Figures 1-1 1-2
SwiGel' Area Resulting From Disintegration Swiaee Area From Successive Cube DivisioJ1.
2-1
Summary of Solids Control Equipment
3-1 3-2 3-3
Derrick Flo-Line Cleaner 2000, 3-Pane! Shale Shaker Motion Shale Shaker Techn%gy & Deck COilfigumfiolls
4-1 4-2 4-3 4-4 4-5 4-6 4-7
Plain Square Weave Rectangular Opening Patented Anti-Blindhrg Derrick Sandwich Screens Derrick Pyramid Screen Positiolls of Solids Oil PWP and PMp.screens Horseslwe Effect Original PMD and Modified Pal/em
5-1 5-2 5-3 5-4 5-5 5-6 5-7 5-8 5-9
Gas Bubble Action in Drilling Fluid Gas Bubble Location & Breaking Action Proper Design of Ml'ld-9a,~:'Separator I Atmospheric Degasser Horizontal Tank Vacuum Degasser Baffle Plates of Hori;.o11lal Tallk VaCiillill Degasser Vertical Tank Vacuum Degasser The Derrick Vacul/m Degasser Flow Diagram for Vacuum Degasser - Uses a Stacked, Corrugated Leaf Design 'Which Alloll's Superior Gas/Liquid Seperation, and Eliminates Remixing and Inefficient Cye/onic Resepel'Cltiolls
6-1
A Typical Balanced Design Hydrocyc1ol1e Operating Properly in . "Spray" Type Undo.flo1\' Solids Discharge A T)7)ial Balanced-Design Hydrocyc1olle that is Ol1.crloaded lI'ilh Solids and "Roping" Hydrocyclonc Unde/,f/ow Adjustnieflfs with Clear :Water Feeding Plugging in Balallced Design Hydrocycloncs Hydrocyclonc 'with Ullde,pow Opening Plugged Suction Screens to Protect Desilters Hydrocyclone Feed Header Problems
6-2
(
6-3 6-4 6-5 6-6 6-7
Figures. continued 6-8 6-9 6-10 7-1 7-2 7-3 7-4
7-5 7-6 7-7 7-8
8-1
Header Problem - SllfIp Discharge Caused by Long Siphon Leg OVCJ,f/OlF Header Problellls Wear Patterns in Hydrocyc/ones Derrick Mud Cleaner Mechanical Solids Control of Weighted Drilling Mud Typical Barite Size Distribution.for Commercially Available Barite Sample Distribution of Drilled Solids in Mud After Passing Through all 80 Mesh Screen Solids Removed by A1ud Cleaner Mud with High % (if Ultra-Fine Solids Should be Centrifuged Bentonite Size Distribution Superimposed Over the Distribution Presented in 200 Mesh Screen Locaf;rJIl of Mud Cleaner OVClf/OW
8-8a-c 8-9a-h 8-10
How Centrifugal Pumps Produce Pressure - Relationship Between Vertical Height (Pressure Head) and Impeller Tip Velocity Work liS. Shut-in Friction Total Dynamic Headior Centr(fugal Pump Centrtfllga/ Pump Cavitation Rig Up Problem - Close Suction Ell Settling of Solids in Pipe Hard Bed Fo1'inatioil - Velocities Above 300ft.l111il1. Will Prevent Settling Friction Loss in Feel o.fHead Pump Curves for Water, Derrick Centrifugal Pumps Pump & Line Problem for Classroom
9-1 9-2 9-3 9-4
Settling o.f Spheres ill Flllids, Stoke's Lall' Settling (~f Irregularly Shaped Particles hi Fluids, Stokes Law Centr(fuge Operafioll Centrifuge Bowll)'pes
10-1
Pro.file View o.f Derrick Hi "G" Dlyer ·with 10 bb/ Sump
11-1 11-2 11-3
LayOllt for Dual Centr(fuge System with Weighted Mud Layoutfor Singe-Stage Centr(fuge System with Weighted Mud Layol/tfor Sil/gle-Stage Celltrigllge Syste11lwitth Unweighed Mud
12-1
An Alternative Method to Detennine Removal Efficiency
8-2 8-3 8-4 8-5 8-6 8-7
Tables 1-1
1-2 1-3 1-4 1-5
Solids Diintegratioll - A Problcm Can Arise WifhOlI{{lI1/l1crease ill Solids Con/ent Micron Si;.e o.fComnuJ/1 Materials API Designationsfor Micron Si;.c Range,'\ Required Force for Stuck Pipe MOl'('menf
1~6 1-7
Solid COlllcllt in Variolls Drilling Fluids Effect qf Chemicals 011 Dispersion of Compacted Miocelle Shale Effects of Common Contaminants on Drilling Fluids
2-1 2-2 2-3
Target Drilled Solids Concentration Optimulll Hydrocyclone Operation Operating Range of CClItr(fuges
4-1 4-2 4-3 4-4 4-5 4-6 4-7 4-8 4-9
Derrick Screen Part Number Composition Speq/ications, Derrick PMD DX Series Spcq'(ications, Derrick PMD HP Series Spec({icatiol1s, Derrick PWP DX Series Spec({icatiol1S, Derrick PWP HP Series Spec(j1cafiolls, Derrick SWG-DX Screens Screen Comparison Chart Areas af PWP and PMD Screens Increases ;n Screen Area for Various Screen Co~nbiJ1atioJ1s Comparative Screen PeJi0I711(fnCe Test Results from Hardin Coullty, Texas Operatiolls
4-10 4-11
i' .
8-1 8-2 8-3 8-4
Example of POllnds Tra/1,~ferrancefor Water I's. Mud Example of Pump Head and System Loss Friction Loss in Fitting - API Units In Approximate Equivalent Feet of Straight Pipe Dynamic Head and Pump Selection, APl Units
10-4
Mud Cleaner Ecollomic Data for FOllr lVells Average Mud Loss by Screen Si;.e Past McAllen Wells Average Cost Well Cost Using Hi "G" Dryer and Cajun Construction
12-1
Solids Analysis Formulas
10-1 10-2
10-3
Formulas I-I 1-2 1-3 1-4 \-5 2-1
Filter Cake Thickness Equation
Required ForIo}" Stuck Pipe Movement Reynolds Number
Drilling Rate in Relatioll10 Reynolds Number Drilling Rate ill Relation to Plastic Viscosity
2-3
Dilution Rate Cost Sewings Calculation Feel of Head
3-1
"g" Faclor
5-1
Boyles Lall' - Pressllre- Volllme Relationship for Gasses
6-1
NOll-Pressured FlolV Through a Pipe
7-\ a 7-1 b
Slliny-Del1sity, Head-Pressure Relationship Feet afHead Requiredfor Proper Hydrocycionc Operatio1l
8-la 8-lb 8-2 8-3
Pressure Head
2-2
9-1 9-2 9-3 9-4
Pressure Head & Equiva/e11l Head Velocity
Centr(fifgal Pump Pressure Head "g" Force for Centrifuge
RPM ~f a Celllriguge Bell Sice Neededfor Cel1lrijilge Stoke's Law
Calculations 7-1 7-2
Low Gravity Solids Concentratioll in Drilling Mild Drilled Solids and Barite Lossfimll Mild Cleaner Screens
12-1
Calcli/ations .lor Solids Removal System
PCliol111GnCe
Effects of Drilled Solids
Chapter 1
introduction Many years ago, a controversy raged about the effects of drilled solids on well costs. Many thought dri11ed solids were an inexpensive substitute for weighting agents. As drilling encountered more difficult environments, however, hole problems finally became undeniably associated with excessive drilled solids.
Most rigs now process all drilling fluid sequentially in accordance with established practices. No matter what mud is used (gel-water, lignosulfonate-treated, lime-based, oil, invert emulsions, dispersed, nondispersed, etc.), drilled solids are undesirable. All drilling fluids must be treated sequentially to remove smaller and smaller drilled solids. Solids control equipment was histotically justified as an insurance policy for catastrophe prevention. Today, the more expensive drilling fluids require a low drilled solids concentration. Polymers that adhere to active solids are expensive, and these drilling fluids requlre lower concentrations of drilled solids than Iignosulfate drilling fluids. Todays environmental concerns dictate the removal of drilled solids to minimize wastes. Economics also playa part -- expensive drilling fluids can be reused if solids are removed in an efficient manner. And if wastes are minimized, so me disposal costs.
In summary, the reasons to eliminate drilled solids me many:
• • • • • •
To reduce wastes for c11I'ironll1enta/ reasons; To save money OJ1 drilling fluid costs; To save money
011
To prel1ent blo)!,
waste disposal costs;
ollIS;
To prevent stllck pipe; and To prevent excessive !Vear 011 drill bits.
The surface mud system is a processing plant to prepare mud for return to the bottom of the hole. Since all wellbores are not the same, some condition.s require more rigid mud specifications than others. TIle mud system is usually the key to drilling the least expensive footage. If the mud is impropedy treated, trouble costs can skyrocket. Most of the chapters in this manual will relate to the removal or elimination of drilled solids and other undesirable contaminants. Some of the effects of drilled solids will be discussed in the section that follows.
1-1
Chapter 1
Effects of Drilled Solids
Effects of Solids on Viscosity When solids are added to a water-based mud, some of the free water becomes chemically attached to the solids. This decreases the amount offree liquid and increases the fluid's
vi.o:;cosity. The amount of water absorbed by a given amount of solids is a function of: 1. 2. 3. 4.
The The The The
particle size of the solids; reactivity of the solids; type of drilling fluid; and type and amount of chemical additives present.
This chapter will deal with the effect of each of the above listed variables separately in the order listed, even though the effect on viscosity is the total of all four acting at the same time.
Particle Size The range of particle sizes taken into the mud stream at the bit will depend mainly on: Formation hardness; Bit type; Chip hold down; and Effectiveness of the hydraulic hole cleaning action. Even if adequate hydraulics are being used and chip hold down or differential pressure is normal, in hard formations fine chips will usually be generated. Most ofthe particles are finer than 22 microns. Another reason for fine chips is the bit type -- a diamond bil will generate smaller chips than a rock bit in hard formations. Use of a water base mud will result in smaller chips than oill11ud because of hydration and dispersion effects of the aqueous system. Use of a water base mud for drilling a soft shale fomlation will ordinarily result in very fine cuttings. With an oil mud and a cone bit, most of the cuttings are larger than 30 microns. Use of a PDC bit will produce cuttings which are typically much larger. TIle pm1ic1e size distribution is also affected by mechanical degradation. As the particles are circulated up by the mud return system, they are mashed and ground. The corners me broken off, so that continually smaller and smaller particles are being generated. Borehole stability is a factor affecting the mechanical degradation of formation solids brought to the surface by the mud system. If the hole is enlarged by erosion or sloughing, cuttings cannot be lifted from the hole efficiently. Low annular velocity in the enlarged sections of hole will cause cuttings to fall back and be ground to a finer size before reaching the surface.
1-2
Effects of Drilled Solids
Chapter 1
Figure 1-1 SU1:face Area Reslllting From Disintegration
Volume = 1".1
Surface Area = 6"
"",
l=t'",'
:!
1"
Divide Into Eight Y2" Cubes
Volume 1"
-~
Surface area 12":!
~ ~ ~ f+"',,,',,,,",
- "," "'-.<»
R=- 4f:f 4-
Divide each Y2" Cube into Eight 14" Cubes,:= 64, tA" Cubes Volume is still I" ,1 Surface Area = 24" 1
I I I I I I I I I I I I I I I I I I I I I
[
[
I I I I
[
I
[
Volume Remains Constant at 1":; Smface Area Increased 600%
Liquid Volume Adsorbed Increased 600%
1-3
I I I I
Effects of Drilled Solids
Chapter 1
Drilled clillings should be examined closely. If they have sharp angular edges, the hole is being cleared properly. If the edges arc smooth and rounded, the hole is not being cleaned properly and a change in the rheological properties or the annular velocity of the mud. or both, is in order.
In addition to the mechanical degradation of the drilled solids, chemical actions also calise them to disperse or break lip into smaller and smaller pieces.
A Solids Problem Can Occur Without an Increase in Solids Content The surface of the solids in any water-base mud system are water set and tie up some of the available free liquid. Figure 1-1 shows the successive divisions of a one cubic inch cube. As this cube undergoes sliccessive divisions, the surface area increases exponentially. As more and more free water is adsorbed by the solids, viscosity increases dramatically. A simple example of this would be a Clip of wheat placed in a quart of water. The tluid, when stilTed, would move almost the same as before the large grains of wheat were added. The viscosity would be almost unchanged. If the same amount of wheat were ground to flour and then added, the resulting dough would be ulany times thicker than the wheat and water. The total amollnt of solids added in each case is the same, but the effect on viscosity is tremendously different. This explains why the solids in a drilling tluid may appear to be in line and yet the plastic viscosity is still abnormally high. For this reason, the size of the solids in a mud (as well as the quantity of solids) must be carefully evaluated.
Figure 1-2 SlfI:f(/ce Areafrom Successive Cube Division 1400 1200
•
"
~
~ t
WOO
600 800
0
w
<00
'00 #00
'"
#0'
#0<
#08
#06
#10
NUMBER OF TIMES THE CUBE IS DIVIDED
1-4
Effects of Drilled Solids
Chapter 1
Figure 1- 2 shows the increase in surface area of a cube versus the Humber of times it is divided. If the solids content of a drilling tluid is plotted against the plastic viscosity, the same shape of curve results. There is an area illustrated by Points 0 lo 4 where there is enough free liquid not adsorbed by a particle's surface area that viscosity is not noticeably affected, There is also an area, illustrated by Points 6 to 8, where so much free liquid has already been adsorbed that a very small change in solids content results in a Im'ge increase in viscosily. The ideal range to run a mud in is between Points 4 to 6. Below this range, very little benefit to mud properties is derived from the increased cost of reducing the solids. Above this range, small tluclUations in solids radicallY affect the fluid's viscosity. To assign a percent solids to the points labeled 0 to 8 on the graph, the size of the solids must be known. For instance, commercial beritonite is made lip of extremely fine solids. The viscosity increase illustrated by Point 8 on the graph occurs at less than five percent. Barite, on the other hand, can have a fairly large particle size and Point 8 occurs al over 50 percent. It takes over ten times more barite to have the same surface area and tie up the same amollnt of Ii quid as commercial gel. Drill solids range somewhere between these two extremes with gumbo type clay being new' bentonite and sandstone being near barite.
Table 1-1 Solids Disintegration - A Problem Can Arise Without Gn Increase in Solids Content Times Cube Split
Total Pieces
Each Cube Size (In)
0
I
I
I
8
1/2
64
114
2 3 4 5 6 7 8 15
512
1/8
4096 32768 262144 2097152 16772160
1116
35 Trillion
1132 1/64 11128 1/256 1/3278 Colloidal Size <1 Micron
1-5
Total Surface Area (In Square Inches)
6 12 24 48 96 192 384 768 1536 196408
Effects of Drilled Solids
Chapter 1
Table 1-2 Micron Si:.es of Comll1OI1 Malerials COMMON MATERIALS
DIAMETER RANGE
MATERIALS
MICRONS
INCHES
Human Hair
30 - 200
.0018 - 0079
Pollen
10 - 100
.00094 - .00394
Commercial Cement Dust
3 - 100
.00018 - .00394
Milled Kitchen Flour
I - 80
.000394 - .00315
Talcum Powder
5 - 50
.000197 - .00197
Red Blood Corpuscles
7112
.00029
,
Table 1-3
API Designatiolls for Micron Si;.e Ral1ges DIAMETER RANGE
COMMON MATERIALS MATERIALS
MICRONS
API DESIGNATION Intermediate
250
2000
Medium
74
250
Fine
44
74
Ultra Fine Colloidal
2
-
44
Less than 2 Microns
Chapter 1
EfJects of Drilled Solids
To discuss the relative size of drilling fluid particles, we must be familiar with the micron unit. A micron is lJ25,OOO of an inch. Table 1-2 helps to visualize this size range in relation to some things that we are familiar with. Particles smaller than 40 microns cannot even be seen by the human eye but m'e larger than 95 percent of commercial barite. The dust in the air that callses our eyes to blink cannot be fell in our eyes but is ten times too large to significantly impact the viscosity of a drilling fluid. The common API designations for particle size ranges are included in Table 1-3.
It is evident that the particles that severely impact viscosity are less than one micron in size. Almost all of the bentonite particles are in this viscosity producing range. The barite particles are a large enough size that large quantities can be added to a system without seriollsly increasing the viscosity. Shale solids will degrade rapidly into the viscosity, producing a range of particles below 2 microns.
Reactivity of Solids Yield point, the second component of viscosity in a drilling fluid, is a measurement of the electro-chemical or attractive forces in a mud. These forces are a result of negative and positive charges located on or near the particle surfaces. Yield point is a measure of these forces under flow conditions and is dependent upon: The surface properties of the lllud solids; Volume concentration of the solids; The electrical em'ironment of these solids (concentration and types of ions in the fluid phase of the mud). High viscosity, resulting from high yield point or attractive forces, can be caused hy: DliUing highly reactive (hydratable) shales or clays that introduce additional active solids into the mud system; and Water can be used to lower the yield point, but unless the concentration of solids is very high, it is relatively ineffective and can be very expensive. Water alone may undesirably alter other properties of the mud. This is particularly true of weighted muds which water affects by increasing fluid loss and lowering Illud weight (necessitating weighing up again).
Type of Drilling Fluid One of the functions of a drilling fluid is to wall the hole with a thin, hut impermeable filter cake. The size, shape, and the ability of solids to deform under pressure will determine the percent solids that can be compressed into a given area. Small thin, flat particles (bentonite) are more effective than large irregular shaped particles (drilJed solids). The type of solids deposited against the well bore changes the thickness of the filter cake needed to achieve a given fluid loss.
1-7
Effects of Drilled Solids
Chapter 1
The ratio of solids concentration in the filter cake to the solids of the mud greatly effects the filter cake thickness. As the solids content of the mud increases, the thickness of the filter cake increases. This is shown in the cake thickness equation.
Formula 1·1 Filter Cake Thickness Eqllation Vf;;::: Filtrate volume T
Vf
T:::: Thickness of filter cake C~ = Vol. frae. solids in filter cake = Vol. frae. solids in mud m A = Filtration Area
e
This equation shows that a reduction in solids content will cause a slightly greater decrease in cake thickness than a proportional reduction in fluid loss. When filter cake problems occur, both options (or a combination of the two) should be considered. The ratio of solids in the filter cake to solids in the fluid must be high to produce thinner filter cakes for a gi\'en fluid loss, Also, the lower the percent solids in the mud, the fewer solids will be available to be filtered out against the well bore for a given fluid loss, Thick, soft filter cakes cause a number of drilling problems:
• • • • •
InCl'eased torque & drag; Stuck pipe; Formation damage; Problems with evaluation and cementing; and Increased swab & surge pressures.
Torque & Drag There are several factors which affect torque and drag. The type of solids in the filter cake affect torque and drag by affecting the lubricity of the fluid and the filter cake. For example, bentonite would have a low coefficient of friction as compared to drilled solids. Thick filter cakes caused by high drilled solids content will result in a smaller hydraulic diameter and increase drag while not circulating. The filter cake should be as thin as possible and contain a minimum amount of abrasive solids. Generally, the solids should be properly controlled before special lubricant products are added to the system,
1-8
EiJects of Drilled Solids
Chapter 1
Stuck Pipe Differential pressure sticking is a common problem. This problem can be lessened by reducing the differential pressure and reducing the area of pipe in contact with the well bore. It is usually not possible to eliminate differential pressure, but the pipe contact mea can be reduced by a thinner, tougher wall cake. As the pipe conlacts the well bore, a force equal to the differential pressure times the square inches of pipe in contact with the filter cake is developed.
For example, let us assume a 400 psi differential pressure, a 9-7/8 inch hole, and 7-114 inch drill collars which are embedded into the wall cake over a lO-foot section. The calculated horizontal force would be expressed in the following equation.
Formula '-2 Required Force for Stuck Pipe Movement F = (I'd) (A) (12 inches/feet) (L) Where: Horizontal force in pounds Difference between hydrostatic and formation pressures in psi Horizontal mea inches~/inch Length of section ill feet
F Pd A L
The following table shows the results of the above equation for 2/32",1/8", 1I4", 112", and 1". The hOlizontal force required is listed along with the over pull based on.2 and.4 coefficients of friction for mud wall cake. ll1ese values are the normal range for most mud types.
Table 1-4 Required Force for Stuck Pipe Movement 2/32"
1/8"
1/4"
112"
1" 10.90
Square inches per inch of collar length
(A)
2.86
3.71
5.27
7.53
Horizontal force in pounds
(F)
137,280
178.080
252,960
361,440
523,200
Over pull required at .2 coefficient of friction in pounds
27,456
35,616
50,592
72,288
104.640
Over pull required at .4 coefficient of friction in pounds
54,9 J 2
71,232
10,184
144,576
209,280
1-9
c:i:ERRICK ~'\"\'~~"'-' EQUJPMENTCOMfI\NY
Chapter 1
Effects of Drilled Solids
The cake thickness reported on the mud report is for a 30 minute period. During trips, the filter cake builds up for a much longer time peliad. The permeability of formations call be much higher or lower than the permeability of the filter paper used in the mud test. The lable above covers the range of possible cake thickness that might actually be encountered down hole. The range between the two over pulls shows the impact of cake slickness while the difference between the columns shows the etlect of cake thickness, Both values are affected by the type and amount of solids present in the dlilling tluid.
Evaluation & Formation Damage The objective in drilling a well is to produce a commercial quantity of hydrocarbon. It is, therefore, most important that the formation be properly evaluated. ll1is becomes increasingly difficult when excessive filtrate invasion and thick filter cakes are present. Production is also hampered by invasion of the solids which migrate into the productive zones.
Cementing The improper placement of cement behind the casing can frequently be traced to mud problems. If the mud is easier to move than the cement, the cement will satisfactorily displace the mud. Mud mobility is related to the viscosity of the mud under the displacement conditions and the area available fortlow. If the mud is gelled in a narrow region behind off-centered casing cement generally will not displace it. The cement flows more easily through the wide regions, resulting in channeling. Similarly, if there is a thick filter cake, the cement will not displace the mud. Thick filter cakes will also create problems in running the casing to the bottom of the hole. ObviouslY filter cakes should be compressible. After the casing is in place, it is too late to condition the mud. Removing drilled solids or improving the mud properties at this time will nor cause the bad mud or mud cake to leave the naITOw regions behind tlIecasing.
Drilled solids removal should be a continuous operation from spud to TD.
Swab & Surge Pressures The filter cake developed under a static condition is always thicker than under a dynamic condition. An equilibrium is reached between the rate of deposition of cake and the rate of erosion while the pump is on. Therefore, thick filter cakes generally develop while tripping pipe; and this is when most of the hole problems occur. Thick filter cakes result in a smaller hydraulic diameter. This can result in lost circulation and well kicks while tripping pipe in or out of the hole.
1-10
Effects of Drilled Solids
Chapter 1
Effects of Solids on Rate of Penetration When all other parameters are equal, rate of penetration valies with Reynold's Number. The [oInmia for Reynold's Number follows:
Formula 1-3 Reynold's NU1I1ber
RE
Where: K
P V D M
Constant Fluid density Fluid velocity Diameter Fluid viscosity
This shows that Reynold's Number is inversely proportional to viscosity. Reynold's Number decreases as viscosity increases. Bit viscosity is a high shear rate viscosity. The limit to which a fluid will shear thill is dependent upon solids concentration, solid type, and solids size. Plastic viscosity is a function offrictional forces within the fluid and is an indicator of bit viscosity. For most bit conditions, drilling rate will vary to the 045 power of Reynold's Number. Thus, we can state that the drilling rate for an operation will change with the ratio of bit Reynold's Number to the AS power.
Formula 1-4 Drilling Rate in Relation to Reynold's Number D, Where: R! R/ D2 D,
= ReYl10fds Nllmber/or Kil'en mild 1I'ilh constant properties = Desired ReYl1ofd:~ Number = Drillil1g mte for Re, = Drillil/g rate for Re,
The plastic viscosity of the drilling fluid is somewhat representative of a high shear rate viscosity; hence, using plastic viscosity as bit viscosity, we can state that: I-II
Effects of Drilled Solids
Chapter 1
Formula 1-5 Drilling Rate in Relation to Plastic Viscosity
D,
' E
PV, \ D, ~
A5
,
Where:
PV j and PY ~ are the respective plastic viscosities cOiTesponding to the drilling rates D j and D~. For example, a fluid having a plastic viscosity of 32 cps produces a drilling rate of 20 feetlhour. When all other parameters are held constant and rock strength is overcome, and plastic viscosity is reduced to 8 cps, the drilling rate could be as follows:
D,
20 (3218)" 20(4)'5 20 (1.866) 37 ft/hour
To drill a 2,000 foot section of hole, the drilling time would be 2,000 feet divided by 20 feet/hour -- or 100 rotating hours. By reducing plastic viscosity and increasing the rate of penetration to 37.0 feet/hour, the resulting drilling time should be 2,000 feet divided
by 36.0 feet/hour -- or 54.05 rotating hours. Plastic viscosity is primmily due to particles under one micron in size. This means that drilling rate is more a function of colloidal solids content than total solids content.
Table 1-5 Solid Content ill Variolls Drilling Fluids
%
lVIud Type
Solids
% Under IlVIicron
%Ovcr 1 Micron
Drilling Rate li'eetffiour
7.2 7.1 2.8 5.9 1.0
23.0 16.5 21.5 22.4 6.0 8.5
A Water
0
B Lab Mud #1
8.6
1.4
7.1 1.8 1l.6 6.1
5.7 5.1
C Polymer Mud #1 D Polymer Mud #3 E Field Dispersed Mud F Lab Dispersed Mud
Effects oj Drilled Solids
Chapter 1
Muds C and D have drilling rates ncal" that of water. even though Mud C has a relatively high solids content. Since particles over one micron have little effect on plastic viscosity. they had lillie effect all (he drilling rate. Muds B, D. and F show drilling rate decreases roughly proportional to their under one micron solids contents. Mud E, compared with F, again shows the slight decrease in drilling rate due to the much higher percentage of large solids. It can, therefore, be concluded that the effect of solids on drilling rate is proportional to their contribution to plastic viscosity rather than their overall concentration. Drilling rate is optimized when plastic viscosities and, hence, particles under one micron are held to the lowest possible level.
Chemical Additives In general, materials added to thin a mud also cause greater dispersion of shale particles. Phosphates tend to be strong dispersants. Dispersion caused by lignites Of lignosulfonates depend upon the alkalinity of the mud. Two percent lignoslilfonate solution at pH 8 causes less dispersion of Miocene shale than distilled waler. At a pH of II, however, the lignoslilfonate solution is highly dispersive. Dispersion of shale pm'tides can be inhibited by certain encapsulating polymers, high concentrations of NaCl, or moderate concentrations of KCL. For practical purposes, chemical dispersion of cuttings is ' eliminated by use of an oil mud.
Table 1-6 Effect of Chemical . . on Dispersion o.fCoJ7lpacled Miocene Shale
Solution
% of Particles Finer than 50 Microns
Distilled Water. . ........ 47 1% NaPO~ ,... ,... ,... ,... ,... ,'" ,... ,... ,... 88 2% Lignosulfonate pH 8,. .. .. 40 2% Lignosulfonate pH 1[ .... ,... ,... ,. , 75 1% NaCl ......................... ............... .55 26% Nael ..................... ". .................. 25
Hole enlargement is usually caused by the hydration of shales. The following are some steps that can be taken to lessen hole enlargement when using water-base muds:
•
Avoid dispersants:
•
Use adsorptive polymers; and
1-13
Effects of Drilled Solids
Chapter 1
•
Use salty mud systems - seawater, NaCl, or KCL depending upon the locality and the nature of the shale. For minimum hole enhu'gement, use an oil mud with Cae!., to give a high salinity for the dispersed water phase (low aqueous activity), For an unweighted mud. the particle size distribution will depend mainly on the factors which control the range ofpartic1e size. Greater chemical dispersion or diamond insert bit drilling will produce a particle size distribution with a concentration of material in the fine end. On the other hand, limited dispersion caused by an inhibited mud or use of mild culler bits in soft rock will produce a distribution of larger cuttings which are easier to remove from the mud system.
For weighted muds. the concentration of barite weight material greatly overshadows the drill solids concentration. API Specifications for barite are as follows:
• • •
Specific gravity = 4.2 min.; Soluble calcium = 250 ppm max.; Ninety-seven percent must pass a 200 mesh screen (74 micron opening).
Various contaminants encountered during dlilling can also negatively affect the chemical properties of the drilling fluid. The following table indicates the effects of some contaminants on drilling fluid properties, such as weight, funnel viscosity, plastic viscosity, yield point, gel strength, fluid loss, pH, etc.
Table 1-7 Effects q{ Common Contaminants Containment
WT
PV
yp GEI.S FL
-
Cement
Gypsum or Anhydrite CaS04 Salt (NnCi) Sulids (Old) Solids (New) Carbonate Bi<.:arbonate Illcr~a~e
FV
~
'" '" '" '" '" '" " '" '" '" " '" " '" '" '" '" '" " '" " '" '" '" " '" "" '" " '" " rknea;~
'"
COIl;lant -
PH
Drilling Fluids
011
PM
PF
MF
'" '" '" '" 01-
01-
01-
CL
-
Ca
Sulids
'" '"
-
-
"
01-
"
-
maybe
"- " "-
-
-
"
-
-
-
-
-
-
-
-
-
-
'" '" " " " "
D~gree
or
-
'"
of
ImTea;~
,. ::.
Decl'ca~c
1-14
Chapter 1
Effects of Drilled Solids
Summary - Benefits of Good Solids Control 1. Increase in drilling rate 2. Longer bit life 3. Better wall cakes: a. Less differential pressure sticking b. Less drag pulling pipe c. Less filtrate entry to fOlTIlations d. Decreased drill string torque e. Reduced surge and swab pressures 4. Lower Mud Cost and Disposal Cost for a. Bm'ite b. Chemicals c. Water d. Cuttings 5. Decreased probability of lost circulation and blowouts 6. Reduced horsepower losses to gel and viscosity 7. Easier to drill gauge hole n. Better hole cleaning b. Better formation evaluation c. Better cement jobs
1-15
Planning
Chapter 2
Planning for Successful Solids Control All drilling programs arc adversely affected by the % drilled solids which remain within
the drilling fluid system. We know that failure to remove these undesirable drilled solids from the mud system can cause slower drilling, the lise of more bits and more rig days on a hole. Excess dilution of the solids increases the amount of drilling waste thus creating excess disposal and mud cost. ll1cse and other factors mentioned earlier add to drilling costs. em-efnI planning of solids removal will make a difference. 111ree common
types of solids control are:
• • •
Gravity or Settling; Dilution; and Solids Removal Equipment.
Gravity or Settling On some land locations. large eruthen pits are lIsed as settling traps, and the pit becomes the solids control equipment. These methods are becoming obsolete and w'e definitely not cost effective.
Dilution Diluting the drilling fluid with water is another means of solids control. As we add water, we increase total volume, which will later have to be disposed of. A small increase of 10% in removal efficiency will decrease dilution fluid by 25% as indicated by the graph below. Compare the amount of dilution required at 70% removal efficiency as to 60% removal efliciency.
Table 2-1 Target Drilled Solids Concentration
"S {g ;>
0
"
~
~
Note.
0% 10% 20% 30% 40% 50% 60% 70% 80% 90%
2%
3%
4%
5%
6%
7%
8%
9%
10%
49.0 44.1 39.2 34.3 29.4 24.5 19.6 14.7 9.8 4.9
32.3 29.1 25.8 22.6 19.4 16.1 12.9 9.7 6.4 3.2
24.0 21.6 19.2 16.8 14.4 12.0 9.6 7.2 4.8 2.4
19.0 17.1 15.2 13.3 11.4 9.5 7.6 5.7 3.8 1.9
15.6 14.1 12.5 10.9 9.4 7.8 6.2 4.7 3.1 1.5
13.2 11.9 10.6 9.3 7.9 6.6 5.3 3.9 2.6 1.3
1 I. 5 10.3 9.2 8.0
10.1 9.1 8.0 7.0 6.0 5.0 4.0 3.0 2.0 1.0
9.0 8.1 7.2 6.3 5.4 4.5 3.6 2.7 1.8 0.9
6.9
5.7 4.6 3.4 2.3 1.1
DIVide the warel mage (bhls) b., the calclIlated dill/cd hole I'Ollll/le (hbI5) to
oblain water dilwioll pCI' bbl hole 1'01//l/Ie. Refer hhl.l' of water dillllion to desired % Igs solids COIIlIllIl, alld read across to I'ellloral role.
2-1
Planning
Chapter 2
Solids Removal Equipment The primary types of solids removal equipment include:
• • • •
Shale Shakers; Desanders. Desilters (Hydrocyclones); Mud Cleaners; and Centrifuges.
How mllch can we afford to spend on the control or removal of drilled solids by utilizing the above equipment? When we look at the cost-ta-benefit relationships, we come lip with some interesting conclusions. In most cases, we can save enough on just the costs of dilution mud, additives and disposal to justify the cost of the best solids removal equipment available. In most cases, the cost for removal by dilution is always excessive in comparison. Note that on a second example for a 70% removal rate of dri11ed solids, approximately 5.7 buncls of dilution is required per each barrel of solids drilled to maintain a weight of 9.0 ppg and/or 5% by volume content of drilled solids. Assume a drilling rate of 50 ftlhr and 24 hours of drilling time with a 12-1/4 inch bit. The quantity of drilled solids could be in the range of 175 banels of cuttings. At 70% removal rate, dilution could be in excess of 998 biUTCls.
Formula 2-1 Dilulion Rate bbl dilution
Drilled solids to dilute (bbls) Target dIilled solids concentration, %
For example: bbl solids to dilute
= 175 bbl cuttings x (I
- 70%)
=52.5 bbl Solids
bbl Dilution = (52.5 bbl Solids / .05) - 52.5 bbl Solids = 997.5 bbl, Ot" 5.7 bbl dilution pel' each bbl of hole ddlJed Consider now, if improved solids removal control equipment is utilized and the removal rate goes lip to 80%, then only 3.8 banels of dilution fluid will be required for each bmTel of solids drilled and/or a total of 665 harrels compared to the 70% removal rate of 997 balTels. This could be a savings of 1.9 balTels of dilution fluid for each band of hole drilled. Consider the cost savings, even with a nominal fluid cost of $8.00 per batTel. 2-2
~ '. ...
,'"
'Co:)
1(g' ,,.., .... "" '~
. '"
~
,6>.
. )J~ ~ ... ,.
I~.J
,,~
".2.'
H'tIl""en>"'&"""!' .... eo
«
1~\
_.
I
r' ~ u ...".,.
"1',,,,
IS "',.,..... ,,,,90,
83m.
1«
Ie",.,.,..
~. ~~
FcC 'oW, fl.·~In . PIl"'''
..oce
,,,.,,,,,
._.
FLC ,C.OI.I ..
c", ... "
,~ i
It''':''-.
1 I .~;,l
..
[, :
I'll.
I
l
I
,",,,, < 1I
•
+-~I
'~ " """.
""""
'fl'''''
o"'''"~,,
.,Q",,",,",:!,,,,,,
~_ _._"
o~·,o,o ~. """". ,
I
O"i".,
Ul'"~~_~.""'~
~",;~CMG"",",
~ Size Range of Solids Removal by Equipment Type EqUlcment
5'1pE'nltitn l~art!;:I;a SI:ze
FIG-Line Primer FLo-Une Cleaner 2000 Mud Cleaner
1000 Miaons eo Larger (10 mesh ) 65 Microns
Desar:der Desiller DE-lOCO Centrifuge
40-60 M icrons 20-4 0 Microns
WeiGhted & Close!;! Systems Un...:eigllted Un",eighledAppliC
2-7': Microns
Un'.veishtedIWelgh'ed Closed Systems
5S Microns
",
""
PrimM' Anpl
.,
Gumbo Allplicalio:lsUnweighlod
am«0"" ' '''9 0'
L... 4'.... 'oy,dl
I.""".
Planning
Chapter 2
Formula 2-2 Cost Savings Calculatioll Cost Savings = (997 hhl Dilution - 665 bbl) x $8.00 bhl = $2,656 Savings Also, consider the reduction of any waste disposal expenses that can be included. 1l1is certainly will justify the need for proper solids control equipment based on the
equipment's ability to increase the solids removal efficiency by only 10%. As the cost of the drilling fluids increase, even a small percent in solids removal efficiency can be justified. A typical solids removal process involves passing the mud through a seties of mechanical devices. Each in turn removes a portion of solids, depending on its removal efficiency, the amount and the size of solids in its feed.
Shale Shakers The first point of solids removal is at the flowline. The shale shaker requires less maintenance than other equipment, is the easiest to operate, and can be the most effective phase of the solids removal system. The shaker should be designed to use the finest screens possible with the least loss of whole mud. Screen selection is a compromise between solids removal, circulating rate and the dryness of discharge. The pm1icle size that a shaker will sepmate from the feed pm1icJes is largely determined by the screen opening size. The "New" API Screen Designation (API~RP~13E) should be on all screen panels. API-RP-13E states that all screen panels should be labeled using a permanently attached tag, and should include the following information:
• • • •
Manufacturers designation; D16, D50, & D84 cut points in microns; Screen area available for screening (ie: non blanked screen area)~ and Conductance.
The actual separation size and cutting dryness will be determined by many other factors, including particle shape, fluid viscosity, screell deck angle, vibration frequency and amplitude, solids and liquid feed rates, and particle cohesiveness. Shale shakers have capacity limits. Exceeding a capacity limit means excessive mud will be discharged over the end along with the solids. There are two capacity limits on all solids control equipment:
2-3
Chapter 2
Planl1il1g
l. The solids capacity limit is the maximum amount of solids that a device will remove, or convey. 2. The li(IUid limit is the maximum GPM capacity for various drilling flUIds. For the Gulf Coast Area, especially in high yielding "Gumbo" clays, we recommend the L1sage of Flo-Line Scalpers operating in conjunction with the primary units. Flo-Line Scalpers remove sticky gumbo clays before the mud is processed through the shakers. If gumbo clays are not removed before processing at the shakers, they can severely inhibit the effectiveness of the shakers by "blinding" the screens.
Remember that for gumbo or sticky clay, deck angle, not motion type is the key for effective removal.
Desanders, Desilters (Hydrocyclones) Use ofhigh-peIfonnance shakers reduces the dtilled solids loading on the hydrocyclones. However, during the drilling of large diameter hole, high penetration rates and high now rates (greater than 50 feet/hour), 10" or 12" hydrocyclones are recommended to reduce solids loading on the smaller hydrocyclones. There should be sufficient 10" or 12'" cones to process 110% of the mud circulating flow rale. The underflow from these cones, since it is fairly dry, should go to the cuttings pit. If the underflow is not sufficiently dry, it may be further dewatered by screening or centrifuging. If dewatering desander underflows is inconvenient, an extra shale shaker might be used to negate the need of the desander. This may not be true for gumbO areas. A desander may be utilized with a light weighted drilling fluid « 13.0 ppg) to discard sand size solids. Be careful not to discard too much Barite or liquid mud. The next type o[ solids removal equipment, downstream of the shakers or large hydrocyclones, are usually 4" desilters, which may be followed by 2" or 3" microcones. The 3" hydrocyclones may be used for replacements of the 4" units, based on fine screening «175m) ahead of the cyclones. There should be enough hydrocyclones to process at least 110% of the rig mud circulation rate. If hydrocyclones are sized to process surface hole circulation rates, there will be sufficient hydrocyclone capacity for the remaining hole sections. A hydrocyclone underflow will have a estimated 10% to 25% solids content, which means that over 3 to 7 baITels o[ nuid can be wasted with one barrel of cllttings under the best operating conditions. A 2" or 3" cone is more efficient because offinerparticle size separation but the discharge is wetter. A new 3" cone design removes 50% more solids than a typical 4" hydrocyclone. The price [or improvement in efficiency is again paid in terms of discarding more liquid with the solids. A high per[Olmance shaker may be llsed to de-water hydrocyc1one underflows prior to being fed to a centrifuge.
2-4
Planning
Chapter 2
All hydrocyclones must operate at their specified Feet of Head (Fh) for maximum efficiency, gallons per minUle capacity, and particle size cut points.
Table 2-2 Optimum HydrocycloJ1e Operation
10" Desander: 12" Desander: 4" Desilter: 3" DesHter: 2" Desi1ter:
Fll
D50u
80 75 75 100 120
38 45 24 10 08
450 500 50 35 25
Any variance in Feet of Head (Fh) delivery to the hydrocyclones will affect the peli'ormance. For example:
Formula 2-3 Feel
(~fHead
4" Desilter (16 cones) operating at 25 psi, feed mud at 9.5 ppg 25 psi (.052 x 9.5 ppg) = 50.6I'h, actual [Sq. of (50.6.!75Fh)] x 50 gpm = 41.0 gpm perconc 41 gpm x 16 cones = 656 GPM
Centrifuges Cenhifuges me typically used to:
• • •
De-water hydrocycJone underflow; Remove drilled solids from the active Illud system; and Control rheological properties by removing colloidal particlcs in weighted drilling fluids.
Dewatering For ullweighted drilling fluid. the centrifuge use appears costjustifieu when the drilling fluid and fluid disposal costs exceed approximately $53.00 per barrel. This is based solcly on the economics of dewatering hydrocycIone underflow. As fluid costs increase.
2-5
Planning
Chapter 2
centrifuge use is highly recommended for reducing costs. Greater portions of the circulating flow rate should be processed. The improved separation efficiency that can be derived from wider use of centrifuges is recommended for reducing drilling wastes. Drilled Solids Rell/ova! To minimize the dilution rate of water on a unweighted mud, the centrifuge is cost effective operating on the active mud system. With solids content less than 10%, a centrifuge can operate at high speeds, thus removing a larger volume of the clay size solids. In some cases, this process call be enhanced with the use of a polymer flocculation system. Control (dMlld Properties To minimize drilling waste on weighted muds (OilfWater) two-stage centrifuging is viable provided the centrifuges are properly chosen and adjusted. TIle first centrifuge should be adjusted for solids recovery, with the second centrifuge providing maximum liquid-solids separation. This process also can be enhanced by using a polymer flocculation system on the water base muds only. Proper upstream tlrilled solids removal is necessary to obtain the maximum benefits from centrifuging an active weighted dIilling fluid system. Shale shakers should achieve a solids separation in the 70-75 micron range. Since the majority of the barite disttibution is below this range, maximum dIilled solids can be removed with l1linimal barite loss.
Table 2-3 Operating Ral1ge of Centr(fuges SPEED
G-FORCE
3250
2100
150
2450 2000 1800
1194 1364 1105
150 170 130
717 537 756 756
25-50 10-35 10-35 10-30
UNWEIGHTED: DE-IOOO ( 14"x49") DE-IOOO OB-I (24"x45" ) (24"x3S") OB-2 WEIGHTED: OB-2, OB-3 05-2 05-3
1450 1450 1950 1950
(JS"x2S") ( 14"x30") ( 14"x22")
CAPACiTY
Formula 2-4 G~F()l'cefor
(/ Centl'{fuge
G force for a centrifuge: G's = RPM" x .0000142 x Diameter of Bowl (in inches)
2-6
Plal1l1il1g
Chapter 2
Closed-Loop Systems DIilling wastes are not all hazardous, but we must properly dispose of extra fluids, solids, and other drilling wastes. A closed-loop drilling operation attempts to minimize the quantity of drilling waste and maximize its quality to make subsequent disposal safer, easier. and more cost-effective. A closed-loop solids control system is an essential element of such an operation because it controls both quantity and quality of the drilling waste. Drilling waste, even if not hazardous. may be disposed ofimpropcrly. Worse, it may be mixed with other contaminated waste and become hazardous. Because of the risks inherent with taking waste off location, it is best that waste generated by the drilling operation be processed tlnd recycled back into the environment on location. If at all possible, this should be done in a way that has no negative effect on the environment. In most cases, this is possible with good waste management and the use of existing technology. For several years, the drilling industry has focused much attention on improving drilling waste l11t1nagemenl. The best approach is one that minimizes the quantity of waste, assures that the waste is nonhazardous, and chooses the least costly acceptable disposal method. TIle concept thtlt has evolved is called the closed-loop operation, or system. TIle key to the concept is recognizing that local environmental regulations and individual well ddlling conditions affect the design, implementation and economics of the closedloop system. No single packaged system design or process can be put together and recommended for all cases. Each location has it's own set of variables.
Good planning is the first step to putting together a plan a c1osed- loop system:
closed~loop
system. To
I. Study regultltions in the area and select a safe and cost-effective waste disposal process that is compatible with these local regulations.
2. Identify and isolate all potential waste sources, both on the location and in the sUlTounding environment. 3. Study the overall drilling plan and be sure that all elements of your solids control program are compatible with the expected drilling program. 4. Develop the location layout, grading plan and reserve pit construction to support your closed-loop system. 5. Evaluate the contractor's solids control equipment and steel pits. Use only that portion which will assure maximum dry solids removal and minimum excess fluid waste. 2-7
Chapter 2
Planning
6. Modify and/or add to the contractor's solid removal system as necessary to achieve the most efficient and cost-effective closed-loop solids control system you can put together.
Implementing an Efficient Solids Control System If YOll have done a good job in the planning phase, it is probable that your solids control system willmClke your work easier rather than harder during drilling. However, there
are a few things you should do to be sure you receive maximum benefits from the system. 1. Begin with a proposed solids control program for each drilling intervaL 2. Develop a solids control economic worksheet for each drilling interval. 3. Carefully inspect the solids equipment piping and hook up before spud day. 4. Educate the rig personnel. The best solids equipment in the world is of little value if not run properly. 5. Monitor solids removal effIciency. 6. Maintain peIi'ormance data records on individual solids removal equipment for reference. 7. Follow up on the logistics of disposal to stay ahead of regulatory permitting requirements. This is important to ensure timely perfonnance of disposal operations to prevent drilling delays. Don't wait until your pit is running over, 8. When conditions differ between drilling intervals, a change and/or remTangement of solids equipment may be required.
The above guidelines should help to implement an effective and economical solids control system.
2-8
Chapter 2
Planning
Equipment Operation Guidelines Shale Shakers I. Operate shakers with appropriate mesh screens for solids-laden fluids at maximum flow rates over any significant hole interval wilhout significant fluid loss over the screens. Treat 100% of the mud circulating volume through the available units. DO NOT BYPASS THE SHAKERS' 2. For double-deck shakers, run a coarser screen on top and a finer screen on bottom. The coarser screen should be at least two meshes coarser. Watch for a torn bottom screen. Replace or patch torn screens at once. Cover 75% to 80% of the bottom screen with mud to maximize utilization of the available screen area, Flow back pans are recommended for improving coverage and throughput. 3. For a single-deck shaker with parallel screens, try to run all the same mesh screens. If coarser screens are necessary to prevent mud loss, no more than two meshes should be on the shaker at one time with the finer mesh screen closest to the possum belly. The two meshes should have approximately the same size opening. For example, use a combination of J 00 mesh ( 140u) and 80 mesh (I 78u), not 100 mesh (l40u) and 50 mesh (279u). Cover 75% to 80% of the screen area with mud to properly utilize the screen area. 4. Spray bill'S should only be used with unweighted muds. Use as required (mist only) for sticky clay. Spray bars are not recommended for weighted oil base muds. 5. Do not bypass or operate with torn screens; these are the main causes of plugged hydrocyclones. Use screens with mesh backup so that coarser solids will be screened out when the finer mesh tem·s. 6. For improved screen life with non-pretensioned screens, make sure the components of the screen tensioning system, including any rubber supports, nuts, bolts, springs, etc., are in place and in good shape. Install screens according to the manufacturer's recommended installation procedure. 7. Check for sufficient space to install walkways with handrails around the shaker skid to pennit safe service. The shaker skid should be level. 8. Check for correct direction of rotation of the motor for shakers with one vibrator.
2-9
Chapter 2
Plal1l1ing
Cones - Desander and Desilter 1. Operate cones in 'a spray discharge. Rope discharge will not remove solids as well as spray type discharge. Rope discharge will also contribute to excessive cone wear, 2. Check cones for a properly operating pressure gauge. An operating pressure "Rille (fTJllIlllb" for COITeet feet of head is 4 times the mud weight in PSI on the pressure gauge (for cones that operate at 75 feet of head).
3. Operate cones with proper feet of head. (Collsult YOllr centrifugal pUIIIP nWIJl(!c/cIlirer and solids control compally for proper pUIIIP (llld line sizil1g.) 4. A sufficient number of cones should be used to process all of the circulating rate. It is recommended to have sufficient cone capacity to process I J0% of the circulating rate. 5, Make certain the cones receive mud from one compartment and discharge mud to the next compartment downstream. Have a low equalizer valve opened between the two tanks [or proper back flow.
Mud Cleaners 1. Make certain the mud cleaner desillers receive mud from one compartment and discharge mud into a compartment downstream. No other piece of equipmentlike a degasser - should parallel the mud cleaner. In other words, all mud must enter the downstream compm·tment by being processed through the mud cleaner. An underflow equalizing line should connect the two compartments and mud should tlow from the downstream compartment to the upstream compartment through the equalizing line. 2. Most four-inch (4") desilter cones process 50 GPM of mud at 75 feet of head. A sufficient number of cones should be used to process all the circulating rate. It is recommended to have sufficient cone capacity to process 110% of the circulating rate.
Centrifuges Weighted Water-Base Muds The pU11)ose of the centrifuge as applied to a weighted water-base dliiling mud is to control viscosity. This is accomplished by concentrating the colloidal-size and claysize solids fraction into the effluent (liquid) stream and disposing of this phase. The economic justification of using the centtifuge on this type of system is the simultaneous return of weight material. Using a centrifuge on weighted water-base muds eliminates the expense of throwing away (jetting) whole mud to control viscosity. Water dilution and excess waste volume are minimized.
2-10
Chapter 2
Planning
•
Add enough dilution water to reduce the effluent viscosity to 35-37-sec/qt Marsh funnel viscosity. If the overtlow is 37 scc/qt or less without dilution, none is needed.
•
Operate lhe centrifuge at a slower feed rate for longer periods to maximize efficiency and provide a more homogeneous mud system. It is better to run the centrifuge for eight hours processing 10 gpm than for four hours processing 20 gpIll.
Weighted Oil-Base/Invert Muds Single centrifuges have typically been run to control viscosity only on weighted water based muds. On an oil base or invert mud system, the cost of discarding the fluid phase and chemicals wilh the colloidal solids can be prohibitive.
Rig-up for Dual Stage Centrifuges The dual stage centrifuge arrangement should be run behind finc screen shakers and/or a mud cleaner. Maximum solids removal efficiency is necessary for solids larger than the barite weighting material (approximately 74 micron). The primary centrifuge usually operates at 700-800 G's with a cut point between 1015 microns. The lllud for the primary or "barite recovery" centrifuge is supplied from the active system. The balite slurry from the solids discharge of the primary centrifuge is returned to the active system, The effluent (normally discarded in a weighted waterbase application) is fed to the second unit, operating at 2000-3000 G's with a cUl point of 2-3 microns. The eftluent is returned to the system; solids between the cut points of the two units (2-15 microns) are discarded.
2-1 I
Chapter 2
Planning
2-12
Shale Shakers
Chapter 3
Shale Shakers Figure 3-1 Derrick® FLC 500 Series
Shale shakers are the most important piece of solids removal equipment. In almost all cases, shakers are the cheapest, most effective means of solids control. If the shale shaker is not being used efficiently, the remaining solids control equipment will not perform properly.
Screens The shaker should be designed to use the finest screen possible with the least loss of whole mud. Selection of a screen is a compromise between solids removal, circulating rate and the life expectancy of a screen. New technology has improved screen life by making screens repairable. Generally, two types of repairable screens are available. The first is a perforated metal plate to which one or more layers of fine screen cloth are bonded. The second type is similar to the first except the metal backing plate is cOlTugated to increase the square feet of available screen area. The particle size that a shaker will separate from the feed particles is largely determined by the screen opening size. The actual separation sizes will be determined by many other factors, including particle shape, fluid viscosity, screen deck angle, shape of vibration, vibration frequency and amplitude, solids and liquid feed rates, and particle cohesiveness.
3-1
Shale Shakers
Chapter 3
Figure 3-2 Shale Shaker MOlion
A. Circular Motion
B. Elliptical Vibrator--_+
C. linear
Straight line
When screens are lIsed with a liquid, a high surface tension film will form on the wires. This film reduces the effective opening to some value smaller than the mesh would indicate. This film effect is increased as mud viscosity increases. This effect plus piggybacking (where particles smaller than the screen opening adhere to larger particles that arc removed by the screen) account for the fact that the screens remove some solid pm1icles sl1wller than the mesh size.
3-2
Chapter 3
Shale Shakers
Shale Shaker Design The design performance parameters of a shale shaker may be described by the type of mOlion, its stroke length and its 11)111. The shape and axial direction of the vibration motion along the deck is controlled by the position of the vibrator relative to the deck and the direction of rotation of the vibrator. If the vibrator is mounted close to the screens and center of gravity, the motion is circular, as shown in Figure 3-2 (A). Cuttings travel direction and speed on a horizontal deck depends on the direction ofrotatiol1, the frequency of vibration, and the amplitude of motion. Amplitude of motion is the distance from the mean position of the motion to the point of maximum displacement. For a circular Illotion, the amplitude is the radius of a point on the screen deck side. The stroke is the total movement or twice the amplitude. If a vibrator is mounted above the deck, the motion is elliptical at the ends of the deck and circular below the vibrator, Figure 3-2 (B). The rate of travel of the solids is controlled by the axis of the ellipse, slope of the screens, and direction of rotation. Shale shakers that use an elliptical motion usually have divided decks with screens placed at different slopes in order to provide proper discharge of cuttings, or the screen deck is tilted in a downhill position. Straight line or linear motion is also llsed, as in Figure 3-2 (C). The rate of travel of cuttings depends on the slope of the motion axis, the slope of the screens, the length of stroke, and the vibration frequency. Linear motion machines can be nm uphill, which allows greater deck coverage by the fluid and provides a higher solids load capacity. USll
3-3
Shale Shakers
Chapter 3
Figure 3-3 Shale Shaker Technology & Deck Configurations Mud Box (Back Tank, Possum Belly)
Screening Surface
A.
Horizontal Basket Single Deck, Single Screen
B.
f------------I C.
Single Deck, Double Screen
(--------... .1 E.
Double Deck, Double Screen
1------------1 - - - - =- - - - - - -
Sloping Basket Single Deck, Single Screen
.......... ' D.
.. .......... .
Single Deck, Triple Screen Cascade, Parallel Flow
........ ..
F.
lOr
....
.. . . .
OR
.........
••
-..........
«*="'''"'"
,
DERRICK
~_""
EOUIPME/ITCOMPA/IIY
3-4
Shale Shakers
Chapter 3
One empirical method Llsed to compare screening devices is the "g" factor, which is flptJitiOlt.:d-to-tlinl.';t- \t'"here:,:~- - - - - - - - - - - - - - - - - - - - -
------.,i"t)ligM~I~pl
Formula 3-1 "g" Factor "g" factor = Stroke (inches) x 70490
RPM~
Conventional shale shakers usually provide a "g" factor of less than 3; fine screen shale shakers usually provide a "g" factor of between 4 and 6. The higher the "g" factor, the greater the solids separation possible. The higher the solids capacity, the less tendency there is fOf the screens to blind. Blinding is a term lIsed to describe a reduction of open area in a screening surface caused by plugging or coating. Screens are plugged when solids are wedged or jammed in the screen opening; coating refers to a reduction in the size opening of screen due to a buildup of a film in the wires stich as salt, gypsum, polymer, etc. A "g" factor that is too high for a shaker's screen supports may reduce the life of a screen. Proper screen tension is critical to assure good screen life. Shale shakers have capacity limits. Exceeding a capacity limit means excessive mud wiH be discharged over the ends along with the solids. Capacity limits can only be defined when the screens are not blinded. There are two capacity limits on a shale shaker-. 1. The solids capacity limit is the maximum amount of solids that a device wiII convey. 2. The liquid limit is the maximum gpm capacity for various drilling muds. Usually the solids capacity limit is encountered only when drilling soft, gummy formations or dril1ing large diameter holes with high penetration rates. The overall capacity of the shaker is a combination of the solids capacity limit and the liquid capacity limit.
3-5
Shale Shakers
Chapter 3
Shaker Screen Arrangement If a shnker has lllultijJle screens (one over the top o f anot her), separati on is determined by the finest, or bottom sc reen. The coarsest sc ree n should nlways be the lOp screen. It is very important that the bOHOIl1 screen be checked oflell for tears. (See operationa l g uide lines at the end of this sectio n.) If a shnker has multiple screens in sCli es or if pnrallel shnkers are used, sepnrntioll is delennined by the coarsest screen. Shakers in parallel should use the same mesh screens. (See operational guidelines at the e nd of this section. Gumbo poses major prob lems for sc reening de vices slich as shale shakers. Gumbo in a water base mud tends to slick 10 the screen and is diffi cult to convey off the scree n. Gumbo will no t convey out of a pool of liquid and most "high-pelfonnancc" scree ning machines will have to be tilted dow n to improve the conveynnce . The shaker should be cleaned and initin ll y lilted dow nhill to 15 degrees, fl ow should Ihen be fed 10 the unit. The deck nn gle should bc tilted downhill until the gumbo moves continuollsly from the feed ulltil it di scharges over the end of the unit. A change in sc ree n mesh may be nccessary. Some ri gs have dev ices, call ed "gum bo busters", built into their fl ow lines to re move the gumbo. To minimi ze solids loading effec ts nnd the effect of g umbo, a "scalping" shaker or rotating be lt s uch as Denick's Flo-LineTh I Scalper may be used. The scalping shaker does nOl need to be n hi gh performance shaker and should be run wi th 8 to 10 mesh sc ree ns, Running fincr scree ns on this shaker will uslIall y require more than one scalping shaker to meet screen nrea demands for fluid throughput. Other dev ices which utili ze a no n- vibrating rotnting mes h belt , includin g DelTick's Flo-Line Scnlper, are useful in remov ing gumbo ahend of high performance shakers.
Screen Blinding Problems Screen blinding is oft en a problem when fin e sands nre drilled. Layered scree ns mounted on four inch and smaller ope ning perforated plates loose some of their deblinding nbility. As a result, when sc reen blinding occurs. the ori gin al sand wic h scr ee ns which ma int ain the deblinding feature may be used, but they may not last as long . For Lhis applicatio n, the sta;ldard layered mesh screen should be ti ghtened down properly and the tension checked IJCli odi ca lly to maximize screen li fe.
Operational Guidelines I. Use enough shakers to prov ide s ufficient non-blanked screen area 10 rUIl 100 Illesh or finer screens. Shaker setup should be sufficien t to process solids- Inden fluid s nt ma ximum flo w rates over an y significant hole interval. If enough shakers are not :wailable, treat 100% o f the mud circ ulating vo lullle through the available units.
DERRICK EOUlPAlENr COMPANY
3-6
Chapter 3
Shale Shakers
2. For double-deck shakers, run a coarser screen on top and a finer screen on bottom. The coarser screen should be at least two meshes coarser. Watch for a torn bottom screen. Replace Of patch torn screens at once. Cover 75% to 80% of the bottom screen with mud to maximize utilization of the available screen ,u'ea. Flow back pans arc recommended for improving coverage and throughput. 3. For a single-deck shaker with parallel screens, try to TUn all the same mesh screens. If coarser screens are necessary to prevent mud loss, 110 more than two meshes
should be on the shaker at one time, with the finer mesh screen closest to the possum belly. The two meshes should have approximately the same size opening. FOI' example, use a combination of 100 mesh (140~!) and 80 mesh ((78ft), not 100 mesh (140ft) and 50 mesh (279ft). Cover 75% to 80% of the screen area with mud to properly utilize the screen surface area. 4. Use spray bars (mist only) as needed for sticky clay, etc. Use spray bars only with unweighted muds. Spray bars are not recommended for weighted or oil based muds. 5. Do not bypass or operate with torn screens; these are the main causes of plugged hydrocyclones. Use screens with mesh backup so that coarser solids will be screened out when the finer mesh tears. 6. For improved screen life with non-pretensioned screens, make sure the components of the screen tensioning system, including any rubber supports, nuts, baits, springs, etc., are in place and in good shape. Install screens according to the manufaclUrer's recommended installation procedure. 7. Check the bearing lubrication according to manufacturer's maintenance schedule. Lubricate and maintain the unit according to manufacturer's instructions. 8. Rig up with sufficient space and walkways with handrails around the shaker skid to permit easy service. TIle shaker skid should be level. 9. Check for con'ect direction of motor rotation for shakers with one vibrator. 10. The flow line should enter at the bottom of the possum belly to prevent solids settling and buildup in the possum helly. If the flow line enters over the top of the possum belly, the flow line should be extended to within 8-10 inches of the bottom of the possum belly. 11. Rig up for equal fluid and solids distribution when more than one shaker is used. A cement bypass is desirable.
Chapter 4
Screens
Screens Historically, the drilling industry has ranked shale shaker scree ns by mesh Coullt, opening size, and percent open area. Th e field operati ng procedure has been to run as fine a mesh screen as possible 011 ull weigilled fluids and as lin e as possib le up to 200 mesh (74 microns) 011 weighted fluids . It is unlikely that one size screen call be used throughout the drilling ofa well. Every piece of so lid s control equipment C(lil be overloaded in one of two ways. I, can have more liquid pumped to it than it ca ll process or it ca n have more tOil S per hour of clIttings fed to it than it can cO ll vey. Both lllud density and lllud viscosity affect the throughput capaci ty of a screen . As these va lues increa se, the throughput of the screen decreases. Th e screen 's mesh is the number of openings per linear in ch, counting from the ce nter of a wire. A mesh count of 50 X 50 indicates a square lllesh having 50 openings per inch in both axis directions . A 60 X 40 mesh indicates a rectangu lar opening having 60 openings pe r inch in olle di rection and 40 open ings per inch in the other. The maximum flow rale ofa screen is directly related 10 the open area that the mud can flow through. For this rea son, wire thickness is very important. Two screens wilh the same openi ng but dilTerent w ire thickness will have different flow capacities. The screen with the smaller wire will have a g reater flow capacity due to the larger opening area . It w ill also have a shorter screen life . The double fold hook strip. or " box binding", is best to bind fine screens . Fine screens must la y flat and maintain complete contact with the machine support bed to prevent flutt er and pre mature breaking. The double fold hook st ri p makes it possible to stret ch and maintain uniform tension over the surface of the entire screen. Single fold hook s trips and spot welded hook strips usuall y have shorter screen li fe becau se the tension on the hook strip is not transferred eve nly to the w ire mesh. Over the past several years, many new screen designs and types have created Illuch confusion in the drilling indus try. This is understa ndable when we consider the variety of screens available. For example, screens cOllle wi th one, two or three layers o f cloth , w ith or without a repai rable plat e, and with or without bonding material. Layered or " sandw ich" sc reens have contri bu ted to this confusion. The sa ndwich screen is a combination of two fin e mesh screen s supported by a cale nd ered backing scree n held together ill such a way that the cross w ires of the baltOIll sc ree n crea te an inte rfere nce in the free ope nings of the upper sc reell . Thi s prevents near s ize or oversize particles from embedding in the top screen. The size separation is thus made by the combination of cross wires from both fine screen surfaces.
4-1
5,'creens
Chapter 4
Screen Design Most shale shaker screens and all fine mesh screens are manufactured with a backing cloth. The backing cloth gives support to the top sizing screen to prevent sagging and premature screen failure. All high quality screens are manufactured with a calendared backing cloth. A calendered backing cloth has all of the wire intersection high points tlatlened by passing the \Vire cloth between heavy rollers, giving the material a smooth surface. Backing cloth is calendered to reduce wear on the tine mesh sizing cloth. Generally, two types of bonded screens are available. The first is a perforated metal plate to which one or more layers offllle screen cloth are bonded. A second type is a screen or screens bonded to the backing doth with either adhesive or heat-set plastics. This second type has poorer screen life and has been replaced by the more durable perrorated metal backed screens, The perforated metal backed screens exhibit a vastly improved life span, in part because tears are contained and can be easily plugged or repaired with quick-dlying adhesives. However, a reduction in throughput is realized because of the blanked-off bonded area. further, the nonplugging characteristic of the layered screens is partially lost. In summary, bonded screens have resulted in vastly improved screen life, but with a sacrifice in throughput and plugging resistance.
Screen Weaves Plain S'qllare Weal'e The plain square weave design provides a straight through flow path with the same diameter warp and shute wires in an over-and-under pattern. This is the most common weave. See Figure 4-\ Cor top and side views of the plain square weave design.
Rectangular Opening The rectangular opening design provides maximum open area and tends to prevent blinding or dogging of material. The longer openings prevent material build-up and the smaller dimension controls the sizing See Figure 4-2, for top and side views of the rectangular opening design.
Palented Anti-Blinding
Del'l'ic/cI~
S'Cll1dwich .','creens
This revolutionary Derrick invention prevents blinding without increasing vertical vibratory forces which could work-harden screen wires and cause premature breakage. The Derrick sandwich screens are supp0i1ed by a backing wire screen, which creates
4-2
Screens
Chapter 4
Figure 4-1 Pla;n Square Weal'e
I I
-
I I
..LL
L rr-
--
-
I
I
-
I
I
j
I I
I
TT
If--
Figure 4-2 Rectanglllar Opening
r--r---
I
I
I 1
-
I-
I-
-
II-
I L I I
4-3
I~
Screens
Chapter 4
Figure 4-3 Palen/ed An/i-Blinding Del'l'ick!{' .)'atlc/IV/ell 5/creens
'-
ll-
ff-
I
ll-
L I I I
I
ff-
-
-
an interference in the free openings of the upper screen. This prevents near size to oversize particles {i'om embedding in the top screen. The size separation is then made by the combination of cross wires.
Derrick' Screens This section gives an in~depth look at Derrick screen technology. DelTick Manufacturing Corporation is one afthe world's leading manutacturers of vibrating screens. This technolol:,'Y is the result of over 40 years of experience in the drilling, mining and chemical industries. Derrick 'g extensive research and development in screen technology has produced a major breakthrough in Derrick 'g Pyramid and Pyramid Plus corrugated screen designs. The following topics will be discussed' ' Screen Panels Usage of Screen Cloths & Panels Screen Part Numbers and Ordering Shaker Screen Specifications (API RP 13-E) Screen Installation Screen Selection Gumbo and Sticky Clay Problems
4-4
Chapter 4
Screens
Screen Panels Pyramid® Designation: PlVIDHI The Pyramid is a pateJlled J·dimensional screen surface mOllllted to a perforated plate. This conugated design adds 50% more screen area, as compared to a flat screen. Field and lab studies indicate a fluid handling capacity increase 0[80% and greater. The Pyramid screen is repairable with silicone products for extended screen life.
Pyramid® Plus Desiglwfi011: PlVID+DI
Also a patented 3~dimensional screen, the Pyramid Plus has higher peaks (IS'), which allows more sUlface area per screen panel. The Pyramid Plus has 124% greater screen area than a conventional flat screen panel. For optimum peJi"ormance, two (2) P.MD+ are llsed in the pool area (feed and middle screen panel). This arrangement is recommended for exceptionally high hydraulic demands or large hole volume with high rates of penetration. The combination of Pyramid (PMD) and Pyramid Plus (PMD+) is well suited for difficult applications such as PhPa, pseudo oil based muds, and oil 111uds, Note: Pyramid® Screens are not recommended in sticky clays (gumbo) when being lIsed Wi SCl'eell!i' for a primlllJ' shake/:
PWpTM Screen Panel [)esiKllafioJ/; PWpTM
The PWP features three screen layers bonded to a perforated metal plate, which adds support and repairability. This screen is repairable by using the provided stainless steel plugs, or silicone.
SWGTM Screen Panel Desiglmfiol1: SWCHI
Three screen layers bound at each side by a one piece double-folded hook strip, originaHy designed for use on Derrick's® elliptical motion shakers such as the Derrick Sandwich Shaker and the Low Profile. The SWG will tit all of Derrick's equipment, including the linear Illation shakers. However, use of fine .. than 38 lIlesh is not recommended due to limited screen life on the linear motion shake."s.
4-5
Screens
Chapter 4
Usage of Screen Cloths & Panels Derrick's sandwich screens minimize blinding due to plugging by near size pal1icles and provide higher conductance (11011' capacity) than conventional screen cloths. The sandwich screen is typically used with Derrick's Flo~Line Cleaner. The following types of wire cloth are used in Derrick screens·
DC = Derrick coarse wire cloth OF = Derrick tine wire cloth OX = Derrick extra fine wire cloth HP = Derrick high performance wire cloth SCGR = Derrick high tonnage capacity rectangular cloth P = Slot length runs parallel to the hook strip (This is preferred for the oilfield) R = Slot length runs at a right angle to the hook strip.
Types of Screen Panels for DX Screen Cloth (PMDTM, PMD+TM, PWpTM, SWGTM) Four types of screen panels are currently available for use with DX screen cloth. They are the "PMD", "PMO+", "PWp", and the "SWG" sandwich screen panels. The screen is constructed by 'placing two fine mesh screens over a coarse support screen.
Types of Screen Panels for HP Screen Cloth (PMDTM, PMD+ TM, PWPTM) Four types of screen panels are currently available for use with HP screen cloth. They are the "PMD", "PMD+", and the "PWP" screen panels. I--IP screen cloth is designed to give higher pertonnance by utilizing finer diameters of wire.
Screen Part Numbers and Ordering The table that follows lists the:
• • •
Screen pOllel type; runel she; lj'Pe (!f.\'L'reCn cloth;
(/11(/
!v!esh size.
4-6
Screens
Chapter 4
To prevent confusion in ordering, it is always best 10 use the complete part number, which is composed of all or the above.
Example:
Part Number
SWG'" 48 - 3D DX 38 PWpnl 60-30 HP 125
When ordering screens, give the complete part number of the screens.
Shaker Screen Specifications (API RP 13-E) Genuine Derrick screens have all identification tag located along the hook strip. The tD. tags are very helpful when storing or reusing screens. Following is an example of the information included on Derrick's screen identification tag. This plastic LD. lag states the following information:
Manufacturers Designation: Derrick® & Patented PMD 48-30 DX 250 D-50 cut point in microns: D-50/69 D-16 cut point illlllicrolls: 0- 16/49 D-84 cut point in microns: 0-84/84 Conductance in kilodarcies per millimeter: 1.39 Screen area available for screening in square feet 8.3
DERRlCK® BUFFALO, NY U,S.A. D-50/69 D-16/49 D-84/84
CONDUCTANCE 1.39 AREA 8.3 PATENTED - PMD 48-30 DX 250
Table 4-1 Derrick S'crecn Par! Number Composition
Screen Panel GBGTM SWGDI PWpTlII PMDT1'+I PMD+TM
4-7
Panel Size
Screen Cloth
48-30 48-30 60-30 48-30 48-30
HP DX lIP lIP TIP
Equivalent Mesh Size
60 38 150 125 125
ERRICK fQUIPAlfNf COIJPANY
Screens
Chapter 4
Table 4-2 Specifications, Derrick PMD OX Series (48x30) Screen Name PMO PMO PMO PMO PMO
OX50 OX 70 OX84 OX 110 OX 140
PMO DX 175 PMO OX 210 PMO OX250
u.s. Sieve 48 68 78 100 125 156 174 213
Separation Potential 016 050 084 318 220 181 149 120 96 86 69
231 158 127 105 86 70 60 49
389 269 218 184 143 118 104 84
Flow Capacity Cond. Area PMO+ PMO 6.10 11.9 8.3 11.9 4.18 8.3 11.9 3.53 8.3 2.93 2.29 1.77 1.59 1.39
8.3 8.3 8.3 8.3 8.3
11.9 11.9 11.9 11.9 11.9
Table 4-3 Sj)(!c{jicalio}1s, Derrick PA4/J HP S'erfes (48x30)
Saro1
Namo PMO HP45 PMO HP50 PMO HP60 PMO HP 70 PMO HP80 PMO HP 100 PMO HP 125 PMO HP 150 PMO HP 180 PMO HP200 PMO HP 230 PMO HP 265 PMOHP310 PMO HP460
US Sieve
44 50
57 71
77 105 121 147 168 203 230 261 300 357
Separation Potential 016 050 084 362 299 263 208 183 143 124 101 89 76
62
283 234 207 158 145 113 100 79 67 60 52
55
44
72 59
48
38 31
47
41
388 313 278 221 192 154 133 113 94
82
53
Flow Capacity Cond Area PMO PMO+ 9.51 11.9 8.3 8.20 8.3 11.9 6.78 8.3 11.9 11.9 4.81 8.3 3.93 11.9 8.3 3.20 8.3 11.9 2.59 11.9 8.3 11.9 2.24 8.3 11.9 1.82 8.3 11.9 1.59 8.3 11.9 1.31 8.3 11.9 0.97 8.3 11.9 0.85 8.3 11.9 0.60 8.3
4-&
Screen.\'
Chapter 4
Table 4-4 Speciffcations, Derrick PWP OX Series (48x30) Screen Name
US
Separation Potential
Flow Capacity
Sieve
050.
016
084
Cond.
PWPOX5o.
48
318
231
389
6.10.
5.30.
PWP OX7o.
68
220.
158
269
4.18
5.30.
PWPOX84
78
181
127
218
3.53
5.30.
PWPOX11o.
10.0.
149
10.5
184
2.93
5.30.
PWP OX 140.
125
120.
86
143
2.29
5.30.
PWP OX 175
156
96
70.
118
1.77
5.30.
PWP OX21o.
174
86
60.
10.4
1.59
5.30.
PWP OX 250.
213
69
49
84
1.39
5.30.
Area
Table 4-5 Specifications, Derrick PWP HP Series (48x30)
4-9
Screen
U.S.
Name
Sieve
Separation Potential 050.
016
Flow Capacity
084
Cond.
Area
PWPHP45
44
362
283
388
9.51
5.30.
PWPHP5o.
50
299
234
313
8.2
5.30.
PWPHP6o.
57
263
207
278
6.78
5.30.
PWPHP7o.
71
208
158
221
4.81
5.30.
PWPHP8o.
77
186
145
192
3.93
5.30.
PWPHP1o.o.
105
143
113
154
3.2
5.30.
PWPHP125
100
133
2.59
5.30. 5.30.
121
124
PWPHP15o.
147
10.1
79
113
2.24
PWPHP18o.
168
89
67
94
1.82
5.30.
PWPHP2o.o.
203
76
60
92
1.59
5.30.
PWPHP23o.
230
92
52
72
1.31
5.30
55
44
59
0..97
5.30.
PWPHP265
261
PWPHP31o.
300
48
38
53
0..85
5.30.
PWPHP46o.
357
41
31
47
0.6
5.30
Screens
Chapter 4
Table 4-6 Specifications, Derrick SWG -OX Screens (48x30)
Screen Narr-e
US
Separation Potential
SWGOX50
Sieve <18
050 318
016 231
064
SWGOX70
68
220
158
SWGOX84
78
182
129
388 269 223
SWGOX110
100
149
105
SWG OX 140
125
120
86
SWGOX175
156
96
SWGOX210
174
SWGOX250
213
Flow Capacity Cond. 742
Area 760
4.75
7.60
3.93
760
164
3.21
760
143
2.<18
7.60
70
118
1.94
7.60
86
60
104
173
760
69
49
84
1.50
7.60
All oflheabove conronn IcAPI· RP 13 E. Gut points are listed in microns_ Conductance is lisle:! in k-ioclarcies per millimeters_ Area is listed in square feet of screen area available for screening
·PMD, PMD+, PWp, SWG, OX, PBPand HPare trademarks ofDetTick®Corporation.
.·10
Screens
Chapter 4
Separation Potential (API RP 13-E) The 3 values in the screen specifications denote the screen's separation potential. In order, they are the dSO, dl6 and d84 spherical diameters corresponding to the ellipsoidal volume distribution of the screen 'g openings. The separation potential of a given screen is assumed to be directly related to this distribution. This is usually a valid assumption since the relative separation efficiency of shaker screens lllust be related to the size of their openings, all other factors being equal. The term "separation potential" is used because the actual separation efficiency curve determined experimentally in the field or laboratory wil! not necessarily agree with the separation potential curve. Factors such as so!ids~loading, fluid viscosity, shaker dynamics and drill cuttings shape can have a considerable effect on experimental grade efl1ciency results. However, the ranking of each screen's relative separation ability under similar conditions will be in agreement between API's methods and the actual field separation ability. Because separation potential is a fhnction of only the screen composition, it provides a constant scale by which to rank the separation performance of shaker screens.
Shaker Screen DeSignation System The components of the designation system provide a complete description of the screen's performance parameters. The designation system consists of the following elements:
Mmll!factllrer :s Designatioll This is the combination of letters and/or numbers used by the manufacturer to describe the screen. Screen panel type and composition may be encoded within this designation. d50~
Micron size at which 50% oflhat paliicle size is rejected and 50% passes through the screen.
This is the median aperture size, in microns, of the screen on a volume distribution basis. By convention, it is the single point most commonly lIsed to represent a solids removal device's separation efficiency curve. In grade separation efficiency terms the d50 represents the drill cuttings size \vhich has a 50% probability of removal. Because of the imporlance oflhe d50 size, it is listed before the dl6 and d84 sizes.
4-1 1
d16~
Micron size at which 16% of that particle size is rejected and 84% passes through.
d84~
Micron size at which 84% of that particle size is rejected and 16% passes through.
Chapter 4
Screens
These values are llsed to indicate the range of screen aperture sizes present in the screen. They can be significant \vhen the removal offines from an unweighted mud is imp0!1ant, or conversely, in weighted mud applications where the unwanted removal of barite is a concern. A triple-layer screen will generally have a larger spread between the diG and d84 than a single mesh screen with the same d50 size. Flow ('opacify
Ultimately, the rate at which a shaker can process mud and solids is a function of many factors including shaker configuration, motion variables, lllud rheology, solids loading and plugging by near-size particles. This is an extremely complex system for which no satisfactory model currently exists. However, for a given shaker under constant conditions, flow capacity is largely controlled by the screen parameters of conductance and area.
( 'onductallce The conductance term, in units of kilo darcie simi !Ii meter, defines the ease with which fluid can now through the screen per unit area. In simplistic terms, it can be thought of as the permeability orthe screen with its thickness taken into account. Conductance can also be described as "resistance to flow", and is calculated from the mesh count and wire diameters of the screen. Conductance does not take into account oversize solids occupying screen space or "screen blinding". Hole size and ROP playas large a role in screen sizing as fluid volume.
Area This is the net unblocked area available in the screen panel which will permit the passage of nuid. The area is reported in units of square feet. Some metal-backed (peIforated plate) designs may eliminate as much as 40% of the gross screen panel area due to backing plate and bonding material. Net area must therefore be considered when determining the flow capacity of shaker screens.
Particle Size Measurements The term "particle size" can be misleading because it fails to account for the shape of a particle. That is, all real particles are not spheres and actually have more than one "size". All current solids control equipment is designed to separate pal1icles with the largest dimension at least 5 microns in size. We usually ignore the shape of a particle and use a single size designation, sometimes called the "Equivalent Spherical Diameter". Treating particles as if they were spheres works reasonably well in many cases, but can cause problems when "extreme" shapes
Chapter 4
Screens
such as "slivers" and "plates" of shale are returned froll] the hole. The standard unit of measurement is the Micron.
I Micron
=
25.4 Microns
=
39.37 millionths o1'al1i11cl1 1 Mil (I thousandth of an inch)
25,400 Microns
=
I inch
Normally, one can see a 40 Micron particle and feel a 20 Micron particle. Shale shakers normally separate particles down to 140 Microns in size. High-Performance shakers and mud cleaners normally separate down to 70 Microns in size.
Screen Installation Three (3) screens are needed for each FLC 2000 JP shaker screen frame, with (4) four screens needed for the Derrick fLC 2000 4P. Eight simple steps should be followed when installing a screen, outlined below.
Derrick Rapid Change Bolt System I, * Clean all solids and excessive mud from the scl,-een channel rubbers, cross and side supports, drawbars, and sides of screen frame. 2.
Place the screen on the bed, hook strips to the sides with the "hook" up.
3.* Be sure the screen is flat across the lip of the cross support. 4,
Slide the drawbars onto the hook strips.
5.
Fit the rapid change bolts through the drawbars on both sides ortlle screen. The rapid change bolts are locked into place by rotating the bolt 90 degrees.
6. * A drawba.· stop is located on the right side only (When looking at the shaker from the discharge end towards the back tank). Tighten this side first. 7.
Bolts should be tightened at the center of the drawbar first and then alternating to the ends.
8.
Tighten all bolts on the len side using the center to end tightening method.
* Vel)' Important
4-13
Screens
Chapter 4
The llut on the rapid change bolt does not have to be removed from the bolt. A threadless area on the bolt where the nut stops is sufficient to allow the bolt to turn 90 degrees, releasing the dra'vvbar and screen.
Derrick Ramp-Lok System I. * Clean all solids and excessive mud from the screen channel rubbers, cross and side supports, drawbars, and sides of screen frame. 2.
Place the screen on the bed, hook strips to the sides with the "hook" lip.
3* Be sure the screen is flat across the lip of the cross support. 4.
Slide the drawbars onto the hook strips.
5.
Fit the draw bolts through the drawbars on both sides of the screen. The draw bolts are locked into place by rotating the ramp-10k 180 degrees.
6.* A drawbar stop is located on the right side only. (When looking at the shaker from the discharge end towards the back tank.) Tighten this side first. 7.
Ramp-Ioks should be tightened at the center of the drawbar first and then altemating to the ends.
8.
Tighten all bolts on the left side using the center to end tightening method.
* Ve/J' Importaut The flex-Ioc nut on the spring housing should not have to be adjusted each time ramp10k assembly is engaged or disengaged -- it has been set at the plant. If the ramp-10k requires adjustment, please refer to your maintenance manual.
Screen Selection Proper screen selection depends on the pump output, mud propel1ies, hole size, rate of penetration, and formation type. Use Pyramid® and Pyramid® IJlus for
Chapter 4
Screens
optimum screen perfonnance and drier cuttings. Ask your Derrick sales and service representative to review your drilling program and make recolllmendations for best screen utilization.
Gumbo and Sticky Clay Problems Screening gumbo poses major problems for screening devices such as shale shakers. Gumbo in a waterbase mud tends to stick to the screen and is difficult to convey off the screen. Gumbo will not convey out of a pool ofliquid and most "high performance" screening machines wil! have to be tilted down to improve the conveyance. The shaker should be cleaned and initially tilted downhill to 5 degrees, flow should then be fed to the unit. The deck angle should be tilted downhill until the gumbo moves continuously fi'om the feed and discharges over the end of the unit. A change in screen mesh may be necessary. Some rigs have devices, called GlImho Blisters, built into their flow lines to remove the gumbo. To minimize solids loading and gumbo effects, a "scalping" shale shaker may also be used. The scalping shaker does 110t need to be a high performance shaker and should be IUn with 14 to 38 mesh screens. Running any finer screens on the scalping shaker will usually require more than one scalping shaker to meet screen area demands for fluid throughput, and may actually hinder (he performance of the fine screen linear motion shal{crs below them. PJ'l'lllllid® scree/IS {Ire nol recommended/or s(·,.eellillg gumbo Illilhotli II "scalpillg" shaka No shaker removes gumbo well, but to get reasonable results the bed has to be tilted downward. This is true for machines using any type of motion (linear, circular, elliptical). A better solution to remove gumbo/sticky clay is Derricl{'s Flo-Line Scalper. Convey it, don't shal{e it, for best results.
Chapter 4
Screens
An Innovative Method of Ranking Shale Shaker Screens by Run it/orrison Del'l'ick® Hqllipmen! ('Oll1pUI1Y HOl/ston, Texas Historically, the drilling industry has ranked shale shaker screens by mesh count, opening size, and percent open area. The t1eld operating procedure has been to run as fine a mesh screen as possible on unweighted fluids and as fine as possible up to 200 mesh (74 microns) on weighted fluids. It is unlikely that one size screen can be used throughout the drilling of a well, Every piece of solids control equipment can be overloaded in one of two ways. It can have more liquid pumped to it than it can process, or it can have more tons per hour of cuttings ted to it than it can convey. Both mud density and lllud viscosity affect the throughput capacity of a screen. As these values increase, the throughput of the screen decreases. The screen's mesh is the number of openings per linear inch, counting from the center ofa wire. A mesh count of 50 ~ 50 indicates a square mesh having 50 openings per inch in both axis directions. A 60 x 40 mesh indicates a rectangular opening having 60 openings per inch in one direction and 40 openings per inch in the other. The maximum flO\v rate of a screen is directly related to the opening area that the mud can flow through. For this reason, wire thickness is very impOliant. Two screens with the same opening but different wire thickness wi!! have different flow capacities. The screen with the smaller wire will have a greater flow capacity due to the larger opening area. It will also have a sholier screen life. The double fold hook strip, or "box binding", is best to bind fine screens. Fine screens must lay tlat and maintain complete contact with the main SUppOli bed to prevent nutter and premature breaking. The double fold hook strip makes it possible to stretch and maintain uniform tension over the surface of the entire screen. Single fold hook strips and spot welded hook strips usually have shOlter screen life because the tension all the hook strip is not transferred evenly to the wire mesh. Over the past several years, many new screen designs and types have created much confusion in the drilling industry. This is understandable when we consider screens are available with one, two, or three layers of cloth and with or without a repairable plate or bonding material. Layered or "sandwich" screens have contributed to the confusion and outdated the API method oflabeling screens. The sandwich screen is a combination of two nne mesh screens sUPP0l1ed by a calendered backing screen, held together in
4-16
Chapter 4
Screens
such a way that the cross wires of the bottom screen create an inted'erence in the fi'ee openings of the upper screen. This prevents near size-oversize patiicles fi'om embedding in the top screen. The size separation is thus made by the combination of cross wires fr0111 both fine screen surfaces. Most shale shaker screens and all fine mesh screens are manufactured with a backing cloth, The backing doth gives suppOtt to the top sizing screen to prevent sagging and premature screen failure. AJI high quality screens are nmnutactured with a calendered backing cloth. A calendered backing cloth has alJ of the wire intersection high points flattened by passing the wire cloth between heavy roBers, giving the material a smooth surface. Backing cloth is calendered to reduce wear on the fine mesh sizing cloth. Generally, two types of bonded screens are available. The first is a pelforated metal plate to which one or 1110re layers affine screen cloth are bonded. The second type is a screen that has the fine screen or screens bonded to the backing cloth with either adhesive or heat -set plastics. Both these types of screens give vastly improved life in part because tears are contained and can be easily plugged or repaired with quick-drying adhesives. A reduction in throughput is realized because of the blanked-offbonded area. FUl1her, the nonplugging characteristic of the layered screens is pa!1ially lost. In summary, these bonded screens have resulted in vastly improved screen life at some sacrifice of throughput and plugging resistance. The types of cloth can also valY in mesh count, wire diameter, opening shape, and material of construction. Amoco Production Research has developed a method of taking the final product and determining the equivalent spherical diameter ofpal1icles that will tit through the openings. AI Cutt of Amoco Research uses an image analyzer system to generate equivalent ellipsoidal volume distribution (J) Cult, AI, API White Paper Presentation (July 19, 1990) curves for shale shaker screens. From these distribution cUlves the API will use the D 16 D50 084 as a pal1 of the new API specifications for screen labeling. This procedure allows all shale shaker screens to be ranked by cut point. But cut point does not take into consideration the liquid flow capacity of the shaker screen. The amount of fluid How through a screen is important in determining shaker performance. The amount of flow a shaker screen can handle ultimately determines the size of the screen that can be run, especially during the upper portion of the well which produces the largest volume of cuttings. A unique practical way to rank screens would be by their ability to process flow. One frequently used indication ofa screens ability to handle flow is the percent open area ofa screen. Open area is a twoRdimensionalmeasurement (length x width) of a three dimensional screen (length x width x thickness). Open area is the percent of the screen area (length x width) available for light to pass through. A screen having 45% open area (and 55% blocked by wire) would pass 45% of the light falling on it. Because light travels in a straight path, open area is a good measure of transmitting light through a screen. However, fluid can travel a "tortuous", winding path through the pores of a screen doth, or a series of
4·17
Chapter 4
Screens
layered screens. Therefore, open area is really not an adequate measure of a screens ability to process fluid. Screen conductance is a much better measure. Screen conductance is a measure of the amount offluid that will pass through a screen, Several methods of determining screen conductance have been developed. Flow through rectangular mesh weaves are treated by Wieghardt (l) and Cornell (2) as being similar to flow around submerged cylinders. Ingmanuson (3) recognized the three-dimensional character of screens and treated them as thin packed beds or a porous media. Armour and Cannon (4) treated the screen as a very thin packed bed where the pressure drop through the screen was considered to be the sum of both viscous and ine11ial resistances and highly turbulent flow through a screen as flow through parallel interconnecting channels of valying cross sections. Hoberock (5) treats screen conductance as a telll1 which describes the permeability per unit thickness of a screen. Hoberock characterizes screen cloth as a thin packed bed having porosity, thickness, average opening size and a wire surface area-to-volume ratio. In summary, we can say conductance is the permeability of the screen cloth divided by the thickness or the cloth, (2) Wieghardt, K.E.G., Aeron Quart 4, 186 (Feb 195J),(J) Cornell, WG Trans. Am. Soc. Mech. Engrs., 80, (1958), (4) Illgmanson, W.L, S.T. Han, H.D. Wilder and W.T. Myers Jr., TAPPI, 44, 47 (1961), (5) Armour, J.C., J.N. Cannon AICI-IE Journal Vo!' 14, No.3 (1968), (6) Hoberock, LL, "Fluid Capacity Limits for Shale Shakers," Transactions, 1981 Drilling Technology Conference, International Association of Drilling Contractors, March 10-12, 1981 Calgary, and is given in kilodarcys per millimeter (KD/MM). Screen permeability is determined by the porosity of the screen and wire sutface area, which causes drag on the nuid and restricts movement of tile l1uid through the screen. Conductance can be calculated n'om knowledge of the weave of the cloth, the mesh count, and the wire diameter. In general (but not always), for a given screen opening size and type of weave, conductance increases as the open area increases and as the wire diameter decreases. Derrick® Manufacturing of Buffalo, New York has also used a method similar to Hoberock's. The following table ranks square mesh market grade screens and Derrick's layered ox screens by Hoberock's method of calculating screen conductance. To put these conductance numbers in perspective, each screens image analyzer 0-50 cut point and average micron opening size are also shown. Several conclusions can be drawn from this information. The tlrst and most important conclusion is that Derrick's layered DX screens have vastly superior flow characteristics when compared to market grade screens with approximately the same cut. In almost every case two to three mesh sizes finer can be run when using Derrick's layered DX screens. Market grade screens average a 4. 5 micron coarser image analyzer 0-50 cut point than the average micron opening size. Derrick's layered ox screens average an 11 micron coarser image analyzer 0-50 cut point than the average micron opening size. Further work should be done to determine screen conductance for all shaker screens. A new screen designation system needs to be devised. This new system should fairly rank all screens based on a screens ability to handle liquid flo\".' and achieve cut point
4-18
Screens
Chapter 4
(I) Cutt, Al, API White Paper Presentation (luI. 19, 1990) (2) Wieghardt, KEG, Aeron Quart 4, 186 (Feb. 1953) (3) Cornell, W.G. Trans. Am. Soc. Mech. Engrs., 80, (1958) (4) Ingmanson, W.L, S.T. Han, H.D. Wilder and W.T. Myers Jr., TAPPI, 44, 47 (1961). (5) Armour, lC., IN. Cannon AICHE Journal Vol. 14, No.3 (J 968) (6) Hoberock, Lt., "Fluid Capacity Limits for Shale Shakers"
Table 4-7 !)'creen Comparison Cha1'(
Cut Point (Micron) Avg. Micron LA. D~SO Opening AMOCO (kd/mm) 200 211 8.21 172 6.57 168 I---ISS 136 5.30 4,28 126 104., ---" 245 234 3.36 98 3.11 87 '---' 2.91 182.5 178 2.82 86 72 140 146 2.19 2.18 68 59 [-----. 2.11 106.9 I 105. 123.8 1.89 U7 1.32 74 75.2 I (13 1.19 I 63.7 I Conductance
~,
~
.,19
I
Screen Designation DX70 OX 84 DX 110 OX 140 MG6D OX 175 MG 80 DX 120 MG 100 DX2S0 MG ISO MGJ:f.(L,_ MG200 MG 250
Screens
Chapter 4 (6) Hoberock, LL., "Fluid Capacity Limits for Shale Shakers"
Three Dimensional Shaker Screens Increase Figure 4-4 Derrick® pyramid® Screen
Shaker Capacity... Reduce Equipment Requirements Throughout the preceding two decades, significant advances in solids control have taken place. The seventies produced the mud cleaner. Linear-motion shakers were introduced in the eighties, establishing a new standard and new awareness for good solids control. Shaker manufacturers adopted linear motion with an uphill screening deck to create a continuous fluid layer over approximately two-thirds of the screen deck. As a result, solids control efficiency was dramatically increased by fine screening at the flow line. Beginning in 1984, linear motion shakers, in combination with plated (PWP style) sandwich screens, established a new bench mark of screening performance. Screening with finer screens (J 50 ~ 200 mesh & finer) not only became possible but were demanded in the weighted hole and slim hole sections. The next ten years produced many manufacturers' versions oflinear motion shakers; ho'wever, there has been little departure from the basic concept developed in 1984. For ten years, with the exception of changes in size and ease of operatioll, linear motion shaker technology has held the status quo.
4-20
Screens
Chapter 4
The design innovation that created a new level of shaker peti'ormance came from new screen technology rather than the shaker itself During 1993, Derrick developed three dimensional screen technology, and it successfully raised the standard for fine screening by two to three mesh sizes. This technological breakthrough in screening was significant in itself and would have made a dramatic impact under any circumstances. But when coupled with innovations in drilling teclmology and practices which were simultaneously occurring, the importance and value of the new screen design was dramatic. The three dimension shaker screen could not have been more timely. As the nineties unfolded, new trends in drilling created tougher demands for fine solids removaL Two factors were especially important: (I) the popularity of more expensive drilling fluids such as PhPa, M"MH, glycol, and pseudo oil base muds, and (2) increased hydraulic demands for bit cleaning, solids transport, and motor rotation. Most industry specialists saw improvements in shaker screen performance as the best solution to meet demanding drilling parameters. Derrick Equipment Company focused on this area with the first "Pyramid®" screen, developed in 1993. The name came fro111 a corrugated screen configuration with its rows of parallel peaks and valleys. A Pyramid screen consists of a standard sandwich construction of two fine~mesh cloths layered with a coarse backing cloth. The three are bonded together, then the "sandwich" is corrugated and bonded to a perforated plate. The resulting corrugations, 0.8" taJl, add 50% more screen area to a standard (PWP) perforated plate panel (See Table 4-8). Based on the added area, developers anticipated a 50% performance gain would be realized with the new Pyramid (PMD) screen. Instead, tests confirmed by actual field use showed capacity gains fro111 80% to 100% and greater. Looking further to find the reasons for the dramatic increase in flmv
Table 4-8 Areas (!f"PWP and Pfi/Il) Screens Screen Type
(PWP) Conventional (PMD) Pyramid
4-21
Screen Area (Ft=')
5.3 8.3
Screens
Chapter 4
Figure 4-5 J\)siliuns (!lSo/ids on jJWP and PAID l)'Cl'eens P..1'J'{lIlIid Scrcef/s fM
GRAVITY FORCES SOLIDS INTO TROUGHS
ii',
." i"V'"
.>',1,<
I 'il"II......... , . ..... j
I
j, . .. I ',1,/,1 j
i·. /i ' , ", I II
•
,
j
I
Flat Screells SOLIDS FORM CONTINUOUS BED, IMPEDING FLUID
capacity, it was found that the unique Pyramid constmction allows gravity to continually force oncoming solids down into the corrugation troughs and away n'om the peaked areas (See Figure 4-5). This means the upper pOltion of each corrugation stays relatively clear of solids, allowing improved liquid throughput and enhanced permeability. It is this preferential displacement solids that accounts for much of the increased flow capacity. It should be pointed out that cuttings size, shape, and composition always playa role in screen selection and use, All too often, shaker screen selection and use is based only on mud weight and flow rate. Hole size, types of formations, and expected ROP are equally impOliant, since oversized solids obstruct useable screen area. The solids, or moving solids bed, become a permeable media. The resistance to liquid passing through the solids becomes the primalY performance variable. This is why the preferential
or
4-22
Screens
Chapter 4
placement of solids on a Pyramid screen and the resulting increase in unobstructed screen area is so important The enhanced permeability of tile firsllwo screen panels submerged in fluid helps explain the gain ill fluid-handling capacity using Pyramid (PMD) and Pyramid Plus (PMD+) screens.
Horseshoe Effect Minimized A second advantage of three dimensional screen technology is minimizing of the "horseshoe effect". Most linear-motion shakers have a crown in the screen hame which provides hoop tensioning to secure the screen panels to the screen tl-ame. The result of this design is a flow pattern which concentrates fluid on the outside edges of the discard screen, creating a "horseshoe effect." (See Figure 4-6).
Figure 4-6 Hurseshoe f.flect
By restricting solids/liquid flow between the Pyramid® corrugations, PMD screens provide more uniform distribution of fluid and solids across the entire screen panel. This greatly reduces the "horseshoe effect" and increases screen area utilization. Lengthening the fluid end point allows an increased pool depth 011 the shaker basket and thus provides for more efficient use of screen area on the discharge panel.
"Pyramid Plus" Emerged From Additional Development Because the greatest advantage of Pyramid screens is achieved with the screen submerged in the pool, research was directed at finding the optimum screen corrugation height. Practical design limits the screen height to that of pool depth. Field tfials were made by
4-23
Chapter 4 bonding 110" of corrugated screen on a 40" wide perforated panel. Data from six months oflab and Held testing indicated optimum performance with 80" ofconugated screen cloth bonded on the 40" wide panel. This design, called "Pyramid Plus," uses a corrugation height of IS' instead of the 0, 81/ originally used. This increases screen area for a single 48 X 30 screen panel to 11.9 square feet, 56% greater than the original Pyramid (PM D) screen and 124% greater than PWP screens. Pyramid Plus (PMD+) screens allow shaker capacity to be increased one to two screen-mesh sizes nner than with original PlvlD screens. Table 4-9 shows screen areas for various combinations of screens.
Bonding Layer Modified for Further Improvement Figure 4-7 OriKina/ PAID anti kJod{fied raffel'll
4-24
Screens
Chapter 4
Table 4-9 Increases ill Screen Area/c)!' VC1I'iolf.\' Screen Comhinaliol1s (SCREEN AREA IN SQUARE FEET) Screen
PWP
#1 #2 #3 Total Fl' % Increase
5.3 5.3 5.3 15.9
PMO 8.3 8.3 8.3 24.9 56.6%
I PMO+ and 2 PMD 11.9 8.3 8.3 28.5
2 PMO+ and I PMD
79.2%
11.9 11.9 8.3 32.1 101.9%
On the original PMD, approximately 30% of the screen area is unusable because of the polypropylene bonding material used to hold the layer sandwich screen together By modifying the shape and location of openings in the polypropylene sheet, 11%
Table 4-10 ('omparath'c /:""creell Pc/forma/Ice (1 PMD+ Iii) feed, wi.h 2 miD's) Screen Style
HI' 150 OX 175 HI' 180 HP 200 DX 210 HP 230
PWP
PMD
PMD+
(gpm)
(gpm)
(gpm)
225 250 140 130
325 370 303 245
450 505 380
122
240 175
105
Mud type: Lignosulfonate
340 335 270 MW~
I'erfol"lnance Gain Pt\ID-f- vs PWP PMD+ vs PMD 100% 38% 102% 36% 171% 25% 162% 39% 175% 40% 157% 54%
104
PV~14YP~4
LGS= 11%
more available screen area is created and screen etliciency is increased accordingly. Figure 4~ 7 shows the modified pattern.
4-25
Screens'
Chapter 4
Field Tests Verify Lab Data Table 4-10 shows results arIab tests with PWP, PlvID, and PMD+ screens. Table 4-11 shows a side-by-side comparison of Pyramid Plus screens on a Derrick FloLine Cleaner verses conventional nat screens on a competitive shaker in field use. The
Table 4-11 Tes! re,wlls./i·o!J1 Hardin
('Ollll~Y,
Texas Opera/ions
......... 12.25"'. ...... _... 9.875" .............. 6.5"' Hole Size .. Mud Wcil.!.hl. ........ 8.9 ppg. .......... 8.9 ppg.. .............. 9.5 ppg Pyramid SlTccn .. _...... *PAlD+ DX ].10 ....... *PMD+ DX 210 ......... *PMD+ DX 250 GPM._ ........ 580, .......... -flO.. ............. 258 Competitive Screen ...... XL 8-1-.. .. ........ XL 140 ................. No Test GPM _. .... _._. <290 gpm . . .... <-1-70 gpm m
...
Nole: (PMD+ indicates "Pyramid Plus") (* indicates "improved" Pyramid) Pyramid screen outperformed the conventional screen by three screen sizes.
Summary Screening finer. .. faster, while drilling, is now achievable with Pyramid screens. Since the cost ofinslalling shakers is a significant factor in rig economics, an additional advantage of Pyramid screens is that fewer shakers and less rig space are required.
Degassing
Chapter 5
Problems With Gas Cut Mud Why do we want (0 remove gas from drilling fluids? There are several problems caused by gas, but the main reason we want to remove gas from drilling tluids is to keep the mud pumps pumping mud. Due to the compressibility of gas a mud pump with a compression ratio of 1.5 will stroke with almost no delivery if the mud weight is cut by 33%. A mud pump with a compression ratio of 2.0 (some triplex pumps) will stroke without pumping mud at all when the mud weight is cut 50%. Many think that pumping gas cut mud down the hole is serious hecause it will reduce the hydrostatic head and even lead to a kick. Any bubbles in the suction pit that can be picked up by the pumps and pumped down the drill pipe will probably be smaller than l/S" in diameter. As the bubbles are pumped down the drill pipe, they are under so much pressure and are compressed so small that it will only slightly affect the mud weight and consequently the hydrostatic pressure.
As an example, if you take a 10 ppg mud that is gas cut 10% to 9 ppg, the volume of all the bubbles in one gallon is 231 in-1 limes 10% = 23. J in,l. According to Boyle's Law:
Formula 5-1 Boyles Law - PresslIre- Volume Relatiol1ship for Gasses
P,
v,
P,
V,
Compressing a gas at a constant temperature to double its pressure causes the gas volume to decrease to one-half its present volume. For a 10 ppg mud gas-cut to 9 ppg (23.1 in,l of gas) at the surface at atmospheric pressure (14.7 psi) it will be 2 atmospheres of pressure at 30.6 ft. The gas volume is now 11.55 in-1 and 9.5 ppg. At 5S.S ft. and 4 atmospheres of pressure the volume is 14 of 23.1 in' and mud weight is 9.75 ppg. At lS00ft. the mud weight is 9.98 ppg which is so little the difference can't be seen on a mud balance. The real problem with gas-cut mud is that il can go unnoticed. If the apparent mud weight is low, barite may be added to increase the mud weight to the desired level. If you have a true mud weight of 12 ppg that is gas cut to 10 ppg and add barite to increase the mud weight to 12 ppg in a 1000 bbl system, it would take approximately 1200 sacks of barite. This would be an unnecessary expense in unneeded materials, but the initial cost of the barite would not be the only problem. The added solids in the
5-1
Chapter 5
Degassil1g
Illud would reduce the penetration rate, bit life and solids removal efficiency. If drilling at or near the fracture gradient of the formation, there is a possibility of fracturing the formation and losing all or partial returns. All of the above slow the drilling process and cost money.
Identification of Gas-Cut Mud How do we know if we have gas-cut mud?
• • • • •
•
Compare mud weight between a conventional mud scale and a pressmized scale. If there is gas present the pressurized scale will have a higher weight; The mud weight increases after adding water; A gas detector warning; An increase in viscosity for no apparent reason; Changes in rheologial or chemical properties of the drilling fluid due to chemical reactions; and Appearance of foam in the mud.
What gasses are typically trapped in mud?
• • • •
Methane; CO,; H 2S; and Trapped air.
How do we get gas-cut mud?
•
Influx of gas to the well bore due to an under balanced situation, such as: Pore pressure greater than hydrostatic pressure; and Failure to keep the hole filled.
•
Drilled gas. When a porous gas bearing formation is drilled, the gas influx rate is dependent on the drilling rate. This gas in unavoidable since it is liberated due to the formation being drilled; and
•
Trapped air due to turbulence and drilling nuid properties. (Water will degas itself, but with a drilling fluid, its increased viscosity will have a greater tendency to trap air.
5-2
Degassil1g
Chapter 5
Removing Gas From the Drilling Fluid What do you do if you haven't installed your degasser yet, or your existing degasser isn't working? First of all, don't worry, there are several methods to help remove the gas bubbles:
•
Large bubhles (114 - J") have enough floating force to rise to the surface and break out; and
•
Small bubbles (1/16" and smaller) are too small to float upward ill a mud. There can be so many small bubbles that the mud weight is substantially cut.
Small bubbles are the problem with gas-cut mud. Several ways to decrease the gas bubbles are:
Figure 5-1 Gas Bubble Action in Drilling Fluid
Usually, Large Bubbles Float Easily
Small Bubbles, 1116'" or Less, Rise Too Slowly
~
0 0
So Must Be Removed by a Oegasser
0 0
--
5-3
--
Ii
___
0
~
..Q.....Q..O---L.,P _
---
Degassi11g
Chapter 5
•
As mud crosses the shakers, a lot of the gas will break out. To increase the release of gas, add water in a fine mist spray;
•
Defoamers can be added to reduce the surface tension of the bubbles, allowing more to break out; and
•
Agitate the pit with an agitator. This rolls the mud, letting the bubbles in the bottom of the pit move to the top and break out.
Figure 5-2 Gas Blibble Location & Breaking Action None of
~~-~~~-~., these
~OO 0 0 0 """'" .-....... ~ 0-0 O 0 0..................
---- .-........-.......
----
bubbles can break, none are on the surface.
Flow and drug causes turbulence, which brings bubbles to the sUiface.
All of these bubbles can break.
Derrick's® corrugated suti'ace increases surface area for a thinner sheet of mud and adds turbulence to improve gas removal.
Degassers Mud-Gas Separator One piece of equipment that is often used to remove gas is a Mud~Gas Separator. It is often called a "poor boy degasser" or "gas buster", This piece of equipment isn't a degasser at all. It is usually built at the rig according to the toolpusher's or drilling superintendent's specifications. Because of this, many are inadequately designed and installed.
5-4
Degassing
Chapter 5
Mud-gas separators arc used to break out the lilrgcr gas bubbles (114 - I") from the mud. This equipment should be Llsed ill conjunction with and not as a replacement for a degasser. In general, a mud-gas separator is designed to:
• • • •
Receive Illud from the well; Separate the large bubbles:
Collecllhe mud in the bottom of the separator and flow 1110 the shakers; ancl Remove the separated gas from the top of the separator to a safe place away
fro111 the rig. A mud-gas separator designed for severe kicks should have all or 1110st of the fol1owing specific design features:
• •
The main vent line should run directly from the choke manifold to the Oaif stack;
•
TIle tank should be large enough in dimneter and length to allow for easy sepm'ation of gas from mud, with an easy flow to the gas outlet;
•
The bottom drain in the separator tank should have a U trap to prevent gas from going to the shakers;
This line should be separate from the mud gas separator. This line should be well anchored to hold it in place while gas is blowing;
Figure 5-3 Proper Desigl1 of Mud-Gas Separator Safety Vent 0"....~C;:~=;:==7 Rupture Disc Mud-Gas_
Separator
....
~ r-.. ~
))))))
5-5 EQUIPMEIIT COMPANY
Degassil1g
Chapter 5
• •
The tank should have a pressure safety vent; and The gas pipe from the top of the tank to the outlet or flair stack should be Jesigned for low back pressure, and be non-clogging.
Figure 5-4 Atmospheric Degasser
Casing
Atmospheric Degasser Developed in the early 1970's, atmospheric degassers use a pump to pump gas-cut mud into a tank, spraying the mud in a thill sheet against the wan of the tank. The impact of the mud on the wall causes a separation of mud and gas. The degassed mud then drains from the spray tank. The valve in the atmospheric degassers is subject to plugging by course materials. Do bypass the shakers or the problem will be worse.
110t
Q .~,., ~--,
"
DERRICK
'
'M
__
EQUIPMENT COMMNY
5-6
Degassing
Chapter 5
A small vacuum blower can be mounted on the vent hood of an atmospheric degasser to enhance the removal of gas.
Vacuum Fill Degassers Horizontal Tank
Figure 5-5 Horiz.ontal Tank Vacuum Degasser
Figure 5-6 Baffle Plates of Hori~oJ1tal Tank Vacllum Degasser
Thin Mud Layer Flowing Down Sloping Baffle Plates
5-7
Degassing
Chapter 5
The original vacuum degasser adapted for use with mud is shown on the previous page. It pulls in gas-cut mud in an upstream tank. and discharges into the next pit downstream, with a high equalizer or weir between the tanks. This unit lIsed a long down-slopping bame plate to spread the mud in a thin layer, 118 ~ 3/8" thick if properly leveled. Most of the bubbles are then near the surface of the mud and are easier to break out. The degassed mud fal1s to the lower trough and discharges through the jet pump. The outlet of the jet pump is below the mud level. A vacuum pump on top of the unit pulls enough gas and air olll of the tank and has a suction pipe to lift mud from the pit to the trough in the upper part of the lank. Most of the time there is not enough gas to supply the vaCllum pump, so a three-way valve is installed to let air in and prevent too much vacuum in the lank. Excessive vacuum would fill the tank above the baffle plates and prevent degassing. The three-way valve controls the mud level in the tank.
Vertical Tank
Figure 5-7 Vertical Tank Vacllum Degasser
~o'1 E [ 0
Another type of vacuum degasser Llses a vertical tank and a cone or multiple cones for thin layers of mud to flow over. The Derrick® Vacu-FloTt.! Degasser uses this design, but adds a valuable feature to the cone -- a corrugated surface, which increases the surface area and allows a thinner layer of mud to flow across the baffles. The corrugations add turbulence to the mud and help bring bubbles lO the surface.
5-8
Degassing
Chapter 5
Vertical tank degassers have the same pit arrangement as the horizontal degassers where mud is suctioned into an upstream pit and discharged into the next pit downstream, with a high weir between the pits. On the Derrick Vacu-Flo Degasser, the vacuum is drawn from the top of the tank, which prevents a gas pocket from forming there.
Figure 5-8 The Derrick Vacuum Degasser -- Uses (l stacked, corrugated leaf design lvhic/1 alloH's superior gas/liquid separation, and eliminates remixing and inefficient cyclonic reseparatiol1s.
The power mud for the degasscr should be drawn from the dcgasscr discharge pit and pumped to the eductor. There should be a high weir between degasser suction and discharge pits.
Vent Line f1V111 Tank fo Disposal Area Vent lines need to be large enough to minimize back pressure. Too small a diameter or too long a pipe can cause the gas pressure to blow the mud seal out of the trap. Or, if a float valve is used, too high a pressure can build up in the separator tank. Drains should be placed in low spots to drain water from vapor released from the mud.
5-9
Degassing
Chapter 5
Offshore rigs often use a vent line attached to the delTick leg, which vents above the crown block. Some'land rigs use long lines on the ground that lead to a flare or a burning pit. Make sure the lines are anchored well enough to be secure during periods of high gas flow. Vent lines should transport the gas enough of a distance from the rig that no hazard exists for the rig personnel. Small amounts of methane can be released without burning. Generally, the lines should go to a flare stack to be burned, especially if H1 S is anticipated.
Figure 5-9 Flm!' Diagram/or Vacuum Degasser
Mud from Flow Line
Gas
t Vacu-Flo Degasser
-------+
t
Educto,
25' Max.
j Solids Discharge
Gas Cut Mud
J
Degassed Mud
o
5-10
Chapter 5
Degassing
Advantages of Derrick® Vacu-FloTM Degasser Design Recent test results by a major operator supported the theory that three key elements are needed for superior dcgasser design. These elements are:
• • •
Vacuum;
Impact/Turbulence; and Maximum Baffle (Leaf) Area.
These are the same three elements used in the Den'ide Vacu-Flo Dcgasser design. The Yacu-Flo Dcgasser utilizes a 5 hp vacuum pump to lift mud into the degassing chamber. where unwanted formation gases are removed and then transported by the vacuum pump. The vacuum pump is capable of pulling up to 29" Hg. After the mud enters the degassing chamber, it is forced through circular ports at a high velocity and then impacted against a battle. The resulting impact and turbulence significantly aids in removing the gas bubbles from the mud. The conical baffles (also called leaves) are specially designed to disperse the mud, after impact, over a maximum area. This feature increases the mud surface area that is exposed to vacuum and provides longer retention time in the degassing chamber. The Vacu-Flo Degasser 500 has a total of 5680 " " of leaf area and the Vacll-Flo Degasser 1200 has an increased leaf area of 14528" 2 which is advantageous when operating at higher flow rates. In addition to the elements mentioned above, the degasser incol1)orates design features which enhance performance and operation and provide the Derrick degasser with an edge over its competitors. Some of these features are:
•
Simple Operation. The unit is started by closing two motor switches which energize the eductor supply pump and the vacuum pump. No adjustments are required. Flow rates and vacuums are controlled automatically;
•
Stable Flow Rates. Unlike some competitors which rely on pumps or impellers to geL mud into the degasser or accelerate the mud prior to impact, the Vacu-Flo unit is a true vacuum degasser. Mud is moved and accelerated entirely by vacuum. This is advantageous because flow rates of pumps and impellers are severely affected by gas content in the mud. The flow rates induced by vaclium show little or no change as gas content increases;
•
Two lHodel Sizes. The Vacu-F1o Degasser is available in a Vacu-Flo 500 and a Vacu Flo 1200 (1200 GPM). The essential difference between the two is in baffle area. In order to maintain efficient gas removal at high flow rates, the baffle area must be increased. Many competitive degassers have model sizes that process 500 GPM or 1000 GPM without any change in baffle area;
5-1 1
Chapter 5
Degassing
•
Vacuum Operation. Vacuum increases with increased mud weight. 111is increased vacuum improves gas removal efficiency. Unlike pump-in units, low mud tank levels can be maintained without affecting now rate, The vacuum actually increases as the mud level gets lower. The vacuum draws gas from the top of the unit, eliminating the possibility of a pocket of gas remaining in the dcgasser;
•
Baffle Design. The Vacu-Flo bame is specially designed to provide maximum dispersion. The design also prevents any obstruction of the bafHe area by drill solids or other foreign objects;
•
H1S Rated; and
•
Flexibility of Rig-Up. The Vacu-Flo Degasser does not have to be located directly above the sliction compartment.
DERRICK ~ '~'~'<
.«
- "
, / EOUIPMENT COMII4IH
5·12
Chapter 6
Hyd!'ocyc/ones
Hydrocyclones The shale shaker is still the most important piece of equipment for removing solids from drilling fluid. Hydrocyc!ones are important in solids removal fro111unweighted drilling fluid.
Drilling Rate, Screens, and Hydrocyclones New particles of drilled solids are brought to the top orlhe hole each circulation of tile drilling fluid. Theyvary in size fro111 colloidal to the largest that can be lifted and pass through the annulus. Many factors atfect the size of particles reaching the surface, but the two 1110st important factors controlling cuttings size are closely related: the type formation being drilled, and the type bit being used to drill it. Soft, unconsolidated formations usually are dispersed by the bit jets to their grain size, whether clay (or colloidal), silt, sand, or (more rare) gravel. These formations usually are drilled fast if the solids removal equipment can adequately remove the solids driHed. Often they cannot, and solids removal limits the drilling rate. Firmer formations, still soft enough and not too deep for economical drilling with long tooth bits, usually can be drilled at a fairly high penetration rate. These formations usually produce cuttings larger than the API sand test screen (74 microns opening). Particles which have a dimension larger than 1I4 inch are nearly always associated with either "under balanced" drilling, or hole caving, or both (An exception to this would be drilling shales with oit-based mud, using a PDC bit). Caving can also produce fine material. The volume rate of caving (or sloughing) can be higher even than the fastest drilling rates. Removal of the coarsest of these particles must be done by shale shaker screens to prevent plugging of the downstream hydrocyclones, as well as any olher equipment having nozzle openings. Soft, fast-drilling formations are found in many surface areas of the world. The hydrocyciones, in these situations, may perform much of the removal. They can be the most economical method of removing a high volume of fine solids. In medium hard formations that can be drilled at rates approaching 100 feet per hour or faster, the shale shakers usually can remove most of the cuttings, Even so, use ofa good hydrocyclone installation will result in improved hole, lower drilling cost, and less maintenance work needed on the mud system.
6-1
Chapter 6
Hydrocyc!onc,\'
liquid Attached to Removed Solids When solids are removed from a mud, they are surrounded by a film of free liquid. If silt-size solids cannot bring liquid film with them, they will not discharge from the hydrocyclones. A lesser amount ofliquid film is also on the solids removed by shaker screens. Small solids paliicies have more surt:1ce area per pound than large particles. The amount offree liquid extracted fromlhe mud with the paliicles varies with the surface area per pound of solids removed and the viscosity of the liquid-col!oidal pari of the mud. So the removal ofa given volume and size of solids from a high viscosity liquid will include more liquid than will the removal of those same solids from a lower viscosity liquid.
Recirculating Solids Particles When cuttings arrive at the surface for the very first time, they are the largest they will ever be. They should be removed immediately by screens, hydrocyclones, centrifuges, etc. Ifnot, they will be pumped back down the hole, either to increase the thickness of the filter calw, or 10 return to the surface again. The second time the particle comes to the surface, it will have been broken into more particles ofa smaller size and therefore will have more surface area. These smaller paliicies will be more difficult to remove than the original particle would have been, and if they can be removed they will take more liquid with them. As particles get smaller, they usually have more resistance to breaking lip. The most successful solids removal program with minimum mud loss requires
operating the hydrocyclones and the shaker screens at all times the hole is being circulated, and keeping the viscosity of the liquid phase as low as other requirements will pennit.
Principles of Hydrocyclone Design and Operation Hydrocyclones appear to be very simple. The changing of worn parts and adjustments can be done by almost any rig band. All this can lead people to believe that the separation is simple. This is not so. It is necessary to understand the basic principles of hydro cyclone design to operate hydrocyclones el1iciently and economically.
6-2
Chapter 6
Hydrocyc/ol1CS
Physical Appearance and Geometry The feed chamber (See Figure 6-1) oflhe hydrocyc!one is normally cylindricaL The solids discharge end is conicaL The solids dischal"ge (underflow) opening is often called the bottom, since the hydrocyclone is usually operated in this position. However, it makes little difference inside the hydrocyc!one which end is up. Extending inward and also centered in the top of the hydrocyclone is a vortex finder, through which the liquid dischal'ge must overflow. The feed opening (inlet) is at the top end of the feed chamber and is tangent to the feed chamber wall. There are two types of feed openings used for the feed chamber. They are involute and tangental. Involute flow comes frol11 a circular feed inlet and because of the shape of the inlet, the flow will remain turbulent for a longer period of time, which reduces separation efficiency. Tangenlal flow comes f)'om a rectangular feed inlet, which reduces turbulence and results in increased separation efficiency. The feed inlet can be slightly smaller or slightly larger than the underflow opening, but the inlet and the underflow each will be much smaller than the liquid (overflow) opening, All dimensions, ratios and angles oftbe hydrocyclone are design variables. The inside area of a hydrocyclone is subjected to abrasive wear from the material it is expected to remove. It usually is lined with a lUbber-like abrasion resistant rubber, polykinetic urethane, or ceramic material. Some bydrocyclones are completely constructed of urethane, others paltly so.
Operating Principles During operation, a feed slulTY of solids and liquid is fed through the inlet at a high velocity obtained by steady pressure or feed head as recommended by the manufacturer. A centrifugal pump is used to obtain the desired head. The high velocity into the feed chamber creates a spinning velocity and resulting cenfrifugal forces. The vortex finder causes the stream to spiral downward toward the underflow solids discharge. Centrifugal force and inertia caLise the solids to settle outward toward the hydrocyclone wall, in the downward spiraling stream. The solids separate according to the weight and size of the particles. In the density range of solids palticles in drilling fluids, size is of far more influence than the variations in density. Basically, the biggest particles settle quicker. As the cone section narrows, inner layers of the downward spinning liquid turn back toward the overflow because of the increasing centrifilgal forces near the center. In this
6-3
Chapter 6
Hydrocyclones
balanced design cyclone, as the last of the liquid moves 10 the center and back upward toward the overflow, the downward spiraling solids continue out the underflow, unable to turn back because of their greater inertia and high downward velocity.
Spray Underflow Discharge The solids spiraling down the cone are concentrated by outward settling and because oftbe decreasing space available as the cone narrows toward the end. If the cone is of a balanced design, and if there are not too many solids (by volume) in the downward spiraling stream, the spny undernow appears much like in Figure 6-1, except the particles are traveling so fast they sometimes give the appearance of a shower head. The tilm of liquid associated \vill valY from as high as 90% by volume of the total underflow when only velY fine solids are separated, down to approximately 50% when only very coarse sand is being removed. The inside stream moving upward tov·.rard the overnow at high velocity carries air with it by friction. This air moving upward in the vortex is replaced by air entering upward through the underflow opening. See Figure 6-1.
Rope Underflow Discharge (Overload of Underflow Opening) When the solids load becomes greater, as in faster drilling, there may not be room for ail of the downward moving solids to go out the undertlow discharge, as shown in Figure 6-2. The solids slow down in the "dead" area near this exit, and the total percent separated is severely reduced. It is not possible to illustrate the associated underflow liquid in these diagrams. In rope discharge, the liquid volume will be about 50% of the total, or sometimes even less. The solids that canllot exit from the roping undertlow will immediately be swept up the center stream and report to the overflow. These will include nearly all the smaller and lighter solids since only those with the greatest mass will find their way out the underflow. Some larger solids will go out the overflow. The rope underflow slurry contains velY few fine solids. Therefore, there is less surface area per pound of solids in the underflow when compared to the underflow of a cone operating in spray discharge. Less surface area means less liquid film on each pound of underflow solids, so the roping underflow will weigh more, in pounds per gallon or slurry specific gravity than the spray discharge underflow (feed being the same).
Note: Since more nne solids are lost to the overtlow in rope discharge, the "liquid" overflow fr0111 a roping cone also weighs more than the "liquid" overflow discharge
6-4
Hyd,.ocyc!oncs
Chapter 6
Figure 6-1 A
~)ipical
Balanced Design Hydroc:vc/one Operating })ropel'~)J in "5iJl'oy" ljpe Unde/flow Sol;d~' Lhsc/wrge
Top View
Feed Inlet
-+-+++t}--+-A Underflow Opening
Top View Not (Operating)
Overflow Opening (Vortex finder may be same size or smaller)
ex a
ID
010 _ a
Top
Feed Inlet
~
",>
. ':-. ,..
Liquid Discharge Opening (Overflow)
Vortex Finder .0
Feed -:----H
+ - Working (Inside) Diameter
Side View
Solids Discharge Opening (Underflow)
Hydrocyclones
Chapter 6
Figure 6-2 A jj)pical Halanced-DesiRI'I Hydroc:vc/onc lhal is Overlouded with ,)'otitls und "Roping ".
Solids Reporting to Overflow
+- Rapid Cone Wear Concentrated in this Area "Dead" Area - No Rotation - No Wear
Underflow Discharge Opening
---?--
+-
Slow "Falling" Discharge
6-6
Hydrocycloncs
Chapter 6
from a spraying cone. The mud retllmed to the tanlf.:s will contain solids which would have been .'cmoved with the hydrocyclone opel'ating in spray dischal·ge. The overnow will lIot weigh more than the fecd.
Evaluation Tips
•
"Rope" type discharge cannot remove solids as well as spray type discharge, even though the rope underflow is heavier.
•
The overflow will also be heavier when the underflow is "roping". This means an unweighted mud will increase in weight faster with the hydrocyclones in rope underflow than in spray type underflow. If the hydrocyc!ones are permitted to "rope" in an ul1weighted mud, the only remaining way to remove those solids returning to the system through the hydrocyclone overflow is with a centrifuge, or by expensive dilution (discard pal1 of the old mud and make up new mud), This involves more loss of liquid and good colloidal material than with the hydrocydone in spray discharge.
"Rope" type underflow can result in very rapid wear of the hydrocyclone vortex finder, overflow parts, cone liner and rig pump t1uid end palis. In unweighted muds, it also results in unwanted solids in the filter cake, as well as an increase in mud weight.
•
The ability of a specific hydro cyclone to remove solids frol11 a slurry is lessened by: Feed head increasing above the recommended range; Increased funnel viscosity or plastic viscosity of the feed slurry; Solids (cuttings) pal1icie shapes changing to less spherical; Decreasing solids particle sizes; decreasing the underflow opening size; and Overloading the cone with too many solids (% by volume) in the feed.
6-7
Chapter 6
Hydrocyclones
Size & Design of "Desilters" and "Desanders" I-lydrocyclones have been Llsed on drilling fluids, experimentally or routinely, with diameters varying from 20 millimeters (0.8 inches) up to 30 inches. The earliest applications were large hydro cyclones used to reduce API sand content (particles larger than 74 microns), and the term desander became common. No particular performance standards are associated with this commonly used term.
The processing of the mud through more efficient, smaller 4" hydrocyclones was called "desilting". Since no performance standards were set by any responsible industty agency, any hydrocyclone of approximately four-inch diameter, used on drilling mud, has since been called a "desilter". There are six-inch, even 12-inch, "desander" hydrocyclones in the field that remove more solids from each barrel offeed mud than some four-inch "desHter" hydrocyclones, under certain drilling conditions. If the design theories are the same for both, a smaller hydro cyclone will make a better separation, feed mud and feed conditions being the same. On many rigs, installation and mud system operating procedures are so poor that pertormance differences cannot be recognized. The purpose of the section on the mud circulating system, including this section on hydrocyclones, is to improve that situation.
Field Operation Initial Hydrocyclone Adjustment There is a point at the beginning of most dril!ing operations when the pits are fil!ed with clear water. This is the time, before drilling begins, to balance the hydrocyclones, or to adjust them to a "wet balance" operation.
Balancing Start the centrifugal pump feeding the ("desander" or "desilter") hydrocyclones. Open all undertlows wide open. It will be normal for some water to spin out the bottom in a thin shell-like cylinder and form a spray. See Figure 6-3 (a). Most balanced design hydrocyclones in the field have an underflow opening adjustment. You will need to adjust slowly to a smaller size until the water spray becomes a slow drip, as Figure 6-3(b). To do this, you must be at or near the proper operating head for the hydrocyclone. Leave all the hydrocyclones adjusted in this same way, and observe them during drilling. They \vill discharge solids when there are solids in the feed which can be separated. Otherwise, they will go back to dripping, saving mud automatically.
Hydrocyc/oncs
Chapter 6
Figure 6-3 Hydl'Ocyc1017e Undelilml' Adjllstments
with Clear Water
Fee(/;l1f!,
.I, a,) Open (Too Wet)
b.) Adjusted to Slow Drip (Petiect)
c.) Over-Adjusted (Too Dry)
Dry-Bottom Cyclones and "Dry Plugging" If the hydrocyclone bottoms do not discharge dear water when wide open, the cyclone is on the dry side of balance. Whether by poor design, or by over-adjustment, as in Figure 6-3 (c), removed solids mllst climb across a dlY area called a "beach" in order to discharge. Coarse sand can do this, if there is a opening big enough for it to discharge. Fine and medium silt particles (less than 44 microns) usually will not cross the beach. On the contralY, they will stick on the beach and form a velY dense "dry plug" that can become hard and vely difficult to remove from the hydrocyclone. It is even worse than plugging caused by roping. "Dry plugging" is a vely common problem in hydro cyclone operation and usually it is caused by ignorance of how hydrocyclones operate
6-9
Chapter 6
Hydrocyclones
Operating Hydrocyclones on Unweighted Mud It is desirable in un weighted muds to remove solids to the economic limit. The benefits normally save more money than the removal and disposal costs, Removal of the coarsest solids is not expensive and the results are clearly worth the cost. For best results, (lowest weight mud, smallest remaining solids, least abrasion, thinnest filter cake, slickest hole, high ROP, etc.) for the least total cost and trouble, observe the following rules with whatever hydrocyclones you have available.
Operating Tips I. Start the hydrocyclones before the rig pumps are started on the hole; stop the hydro cyclones only after the rig pumps are stopped, 2. Make sure any holes in the shale shaker screens are repaired immediately, before a large amount of solids reach the hydro cyclones and they begin to plug at the teed. 3. Be sure the shale shaker is never by-passed and that the back tank is never dumped into the active system. Check the plumbing layout of the trip tank. 4. If the hydrocyclones overload (due to solids being drilled faster than the cones can discharge the solids), open the underflow openings out beyond the balance point. When drilling has slowed down, tighten them back to the balance point. 5. Be sure all hydrocyclone suction pits are being stirred. This can increase solids removal as much as 100% by the same set of hydrocyclones, under fast drilling conditions. 6. If one hydrocyclone changes it's underflow behaviOl~ indicating a problem, replace the hydrocyclone, or remove the hydrocyclone, blank the feed header and overflow header nipples. (With practice, this can be done by one lllan during a connection on a good installation with properly arranged hydrocyclones and headers.) The problem hydrocyclone can be inspected and put in order, then returned to the header on the next connection. Some rigs keep a complete hydro cyclone ready to exchange so it is all done in one operation, during one drill joint connection. 7. If plugging of the hydrocydone feed inlets becomes a problem, shut down the centrifugal pump, open the feed header, rake out the material found inside and inspect it. This will let you know whether the problem is from bypassing of the shale shaker (most likely), trash thrown in the mud tanks, or from dumping the back tanks into the active system.
6-10
Chapter 6
Hydrocyc/ones
8. If plugging of the hydrocyclone underflow openings is a problem, and if the underflows are v,lide open, then more underflow capacity (more hydrocyclones) is needed. 9
If you are the supervisor in charge of drilling the wel! and are responsible for the total cost of drilling, issue a simple order to the mud supervisor, "No lllud will be discarded from the tanks without my specitlc permission, each time". This wi!! give you the opportunity to discuss solids removal with people who do not understand it and who dump mud in ignorance. Displacement from running casing and pumping the plug down, spotting a special mud for kick-off, stuck pipe, etc., and plugging the hole back, usually are justiflable excuses for moving whole mud out of the active system.
lO.lf the liquid film of mud on the removed underflow solids causes a disposal or expense problem, investigate the use of a Hi "G"TM Dryer, andlor a decanting centrifuge, to save the liquid colloidal phase and put it back into the system, This can reduce hauling and disposal, and can reduce total cost, although mud system weight will run slightly higher, depending upon which method of salvage is used. This is covered more fully in the section on mud cleaners.
Operating Hydrocyclones on Weighted Mud to Reduce the Sand Conlent In weighted muds, considerable sums of money are expended to keep weight up. We can define a weighted mud as one that contains a commercial solid (such as barite) for the purpose of increasing mud density, and much of tile cost of tile mud is usually due to the weight material. Drilled solids should be kept to an economic minimum. The higher mud weights will tolerate fewer low specific gravity solids than the unweighted muds. Trouble will increase if the drilled solids are not minimized. If the API sand content is 1% or more, hole problems can be predicted. The primary justification for removing coarse solids from a weighted mud is to reduce the filter cake thickness and improve the texture. This can reduce differential pressure sticking, ifsuch a tendency is present. An added benefit will be reduced abrasion on pump parts and other equipment. The reduction in abrasion may be noticeable on the bit, resulting in longer life for the bits (especially for those with unsealed bearings). These benefits must be, and usually are, substantial enough to pay for the barite to replace all solids removed. Otherwise, the process is uneconomical and pointless. By API standards, not more than 3% of tile barite can be sand-size (greater than 74 microns). It will help the filter cake just as much to discard that sand-size barite as to throwaway any other sand-size solids. Some types of barite, however, are very soft and grind to a smaller size more quickly than some of tile drilled solids, Frequently,
6-11
Chapter 6
ff.vdrocycl one,,'
one circulation is sufi1cient to decrease the size Of11105t barite to below 74 microns. It will reduce losses to stop hydrocycloning during "weight up". Enough high performance shakers should be used to run 180 to 230 mesh screens on 100% ufthe mud circulating rate. If this is done, mud cleaners may not be needed. A mud cleaner consists of a fine mesh screen located under the hydro cyclones (as discussed in the section on mud cleaners), Mud cleaners wefe designed specifically to eliminate the sand-size solids and return the liquid and silt-size solids (which includes the desirable barite) to the mud. Barite and drilled solids finer than the screen mesh are returned to the mud system. Larger drilled solids and some barite are discarded by the mud cleaner. The solids returned to the system by the hydrocyc1ones and by the screen will be smaller and less harmful than those removed. If desanding or desilting hydro cyclones are used continuously on a weighted mud, the underflow discard should be monitored closely. Procedures for doing tllis are discussed in the mud cleaner section. The prolonged use of hydro cyclones will result in excessive amounts of barite discarded. Areduction in mud weight does not indicate that the barite is being thrown away. For every 100 pounds of drilled solids removed, 80 pounds of barite and 20 pounds of water must be added to restore the mud weight and volume. The barite added will take less space than the drilled solids. If continued drilling with weighted muds is anticipated, the rigs should install the proper equipment. Three types of solids control equipment are available to remove sand~sized and larger particles from weighted muds:
•
Shale shakers
•
Hydl'o(yclones A:fud cleaners
Shale shakers and mud cleaners are discllssed extensively in other sections. These devices will also remove barite as well as drilled solids, and will also return some drilled solids to the mud system. They work primarily on sizes of solids. The benetits of removing coarse solids from the drilling fluid must exceed the costs of the equipment and the barite loss. For weighted muds, the economics generally indicate that solids coarser than 74 microns should be removed, Centrifuges also work on weighted mud systems, but their purpose on weighted muds is to control rheological properties by removing the velY small particles. This has nothing to do with desanding, and should not be confused with it. The section on
6-12
Chapter 6
~vdro[yclones
centrifiJges discusses the different types and their use on weighted and lImveighted
muds.
Maintenance Good hydro cyclones are high performance devices. They require attention and maintenance just like the traveling and crown blocks, rotary table, drawworks, shale shaker, and other rig equipment. A regular maintenance program wil! minimize the effort required for best hydrocyclone operation, and will assure the best results.
Feed Head (Centrifugal Pumps, Suction and Discharge Lines) Hydrocyc1one wear and performance is highly dependent upon feed head, or feed pressure. The head should be checked regularly at one well intervals or everyone or two months. More frequent inspection is not necessary unless there is evidence of trouble, Some manufacturers specify a required feed head for best performance and satisfactory wear; others specify pressure. "Feed head" refers to a pressure at the feed header equal to that hydrostatic pressure at the bottom ofa column of feed mud, the same height as the "head" specified by the manufacturer. Centrifugal pumps will discharge at a constant feed head if nothing changes in the piping or equipment being fed. At a constant feed heaq, the pressure at the feed header must rise as the mud weight increases. The pressure will also decrease if mud weight decreases. A rule of thumb for the field is four times the mud weight should equal p.s.i. on the gauge for a hydrocyclone header that requires 75 ft. of head for proper operation.
Relations with the Manufacturer It is important to read the manufacturer's instructions concerning your equipment. The instructions should include any details peculiar to the product.
If wear 011 some part of your hydrocyclones is excessive, contact the manufacturer. He should be anxious to help. Wear varies with feed head, feed solids concentration and colloidal content, feed particle size, shape and hardness, mud chemistlY and temperature, underflow adjustment and the removal equipment in operation upstream,
Elastomers Various elastomer compounds used for hydro cyclone manufacturing are aOected differently by various drilling fluid additives. There are good reasons to contact your manufacturer or its representative if you are having a problem with your elastomers.
Hydrocyclones
Chapter 6
Some elastomers wear well in many situations including oil, but do not store well under humid conditions. Some store we!! under those conditions but "grow" in dimensions in the presence of oil because they actually absorb oiL Others do extremely well at moderate temperatures but do poorly at high mud temperatures, etc. When those problems present themselves, contact the manufacturer for advice and solutions. Also, your purchasing agent must be kept informed of the iield conditions and problems for which he is responsible.
Figure 6-4 jJlugging il1 Bu/anced Design Hycll'ocyclones
Tilt· smoolh ,-hope ",dldilC_\" ml(l/IOII dlle lu
JJJlld C'''t"rllig
S/rmgllldml'll (lOlL'wdlccl/csull
Iilejeed
",lei.
Coning Due 10 ParLial Plugging of Inlet
b
Flooding Due to Plugging of Inlet-mayor may not occur depending on manifold design and back prcssure from discharge linc
6-14
Chapter 6
Hydrocyc/oncs
Regular Maintenance At or near the end of every drilled hole, or perhaps during the move on shorter jobs, the hydrocyc!ones should be made ready for spudding in on the next hole. This chore consists of taking each hydrocyclone apart, cleaning up and inspecting each part, replacing those parts that will not last to the next casing point, opening the feed header and cleaning it, and reassembling aU hydrocyclones onto the headers. Bottom adjustments should be left loose or wide open, This general inspection may be needed more than once per hole on deep jobs with long location times. Remember to check the manifold for a build lip of solids and foreign particles, and clean if necess31
Emergency Maintenance Regular maintenance and sensible operation will reduce problems, but there will be emergencies in hydrocyclone operation. They must be recognized quickly and promptly and either corrected or neutralized. They can be extremely expensive.
Feed Plugging, Partial The pat1ial plugging of a hydrocyc!one in operation must be corrected immediately, Figure 6-4 shows why. Partial plugging of the feed is caused by hard objects, such as welding rod pieces, oyster shells, large rocks, pieces of cement, etc. in the feed orifice. This causes a reduction in velocity in the inlet, just as if the centrifugal pump did not furnish enough head. The hydro cyclone action is lost, there is no separation, and the hydro cyclone cannot balance. It acts as a swirling funnel, all or Illost of the Illud entering goes out the bottom in an inverted cone shape. Even clean mud from the overflow header may flow backwards into the hydro cyclone and be lost. The mud loss at the underflow may be as high as 50 to 75 barrels per hour, even in a four inch hydrocyclone. A hydrocyclone improperly operating should be removed from the header immediately. ]fthe problem is easily and quickly corrected, the cone can be reinstalled on the header; otherwise, the header stubs must be blanked, or a spare hydrocyclone put in place. With practice, this can be done in the time of a drill pipe connection if the hydrocyclone unit is properly designed and maintained.
Feed Plugging, Complete If a hydrocyclone feed inlet is completely plugged, it is caused by a soft object, such as a rag, glove, hat liner, piece of rubber iTom a cementing float plug, etc. See Figure 6-4 (b). These objects can be difficult to remove. All the mud loss in thiS case is cleaned mud
6-15
Hydrocycloncs
Chapter 6
from the overflow ofthe other hydrocyclones. The mud loss will be extremely high, and floods straight down with no rotation. Again, it cannot be tolerated fi'om an economic, or from an environmental standpoint. Stop the pump, remove and repair or replace the hydrocyclone, or plug the headers, and restart the centrifugal pump as quickly as possible.
Figure 6-5 HydI'OL}'c/one with Undef.11ow Opening Plugged
Coarse Solids Passing on to Rig Pump, Filter Cake, Annulus Severe Wear Inside Vortex Finder
/
t
Ring of Severe Wear
" ' - Still Bed of Very Coarse Material ~ Plugged Area
0-16
Chapter 6
Hydrocyc/ones
Underflow (Solids Discharge) Plugging lfthe underflow is plugged, the cause and result are quite different fium feed plugging. See Figure 6-5. One cause of underflow plugging is overload afsolids in the feed, or too much restriction of tile underflow opening. Usually poking a small blunt rod up the underflmv will start it flowing again. The underflow should be adjusted to a larger opening to obtain spray discharge, if at all possible,
If it is extremely difficult or impossible to poke a stnClll rod up the underflow, one of two things is wrong: 1, The underflow may be adjusted so tight there is no opening and no space for the
rod, or for the solids to flow out 2. There may be a very hard plug of baked fine silt. This is caused by the underflow opening being adjusted smaller than the balance point in a slow drilling situation, usually in a deep hole with hot mud. The fine silt cuttings cannot flow over the dry beach, so they extrude as a plug which then bakes into an extremely hard mOltar. (Refer back to Figure 6-3). Removing this plug can be quite a problem, and is easier if the cone is removed. The adjustments must be opened until the solids can flow. Ifa hydrocyclone is permitted to operate with a plugged undernow, all the cuttings in the feed mud will pass through the hydrocyclone to the overflow and through the rig pump to the hole. The wear from the hydrocyclone oveIilow is velY severe. The wear on the rigs pump expendables is at a maximum, and the cuttings pumped back to the hole will reduce penetration rate, increase bit wear, and will cause sidewall sticking in most areas. A plugged underflow increases the work load on the drilling crew and also jeopardizes the drilled hole. The hydro(l'c/olle JJlllst Hot he permitled to operate with a pll1XXed 1II1dqi7oll'. If cared for all a regular basis, the hydro cyclones can make work easier on the rig. I r not, they will help nothing and will only cause trouble. If the hydrocyclone installation is adequate and properly operated, the hydrocyclones will rarely plug. If the number of hydrocyclones is not adequate, they will be overloaded vvith solids, and they will continually plug. It \~Iill be literally impossible to keep them operating properly. Low silt muds obtained by good desilting practices can reduce rig pump maintenance and reduce differential wall sticking. That alone is worth all the effort of caring for the hydro cyclones.
0"17
Chapter 6
Hydrocyclones
Hydrocyclone Sizing Determine Minimum Hydrocyclone Feed Capacity The minimum hydrocydone feed capacity required depends L1pon the total flow into the hydrocyclone suction compartment This flow can be much larger than the normal rig circulation rate. This now rate includes tvlaximum rig circulating rate;
• •
The mud gun flow upstream from desilter discharge
•
Any other flow entering into the desilter pump suction compal1ment, or into any compartment upstream of the desilter suction, from any compartment downstream of the desilter suction compartment. (Flow should not include the volume of Illud being processed by any removal device upstream from the desilter ifit is installed as recommended). Please see section on mud system arrangements for cost savings ideas for mud system set up.
compartment~
The degasser eductor pump flow if the suction is downstream of the hydrocyclones; and
Install enough hydrocyclones of each t)'pe to process at least 110°;;) of the total calculated flow rate. Each hydro cyclone unit should have its own centl'ifugall1ump!
Special Problems In certain situations, factors other than pump rate may require a set of larger hydrocyclones ahead of the desilters. Some rigs seem to work ill situations, or with personnel, that prevent "clean" mud systems. Plugging or hydrocyc!ones may be a constant problem. In these cases, suction screens may be helpful but the best solution is often high performance shakers and good operating practices.
Agitation In the removal system, proper agitation will provide a uniform feed to the hydrocyc!ones, and should increase the removal of solids. Mechanical agitation should be used if at all possible.
6-18
Hydrocyc/ones
Chapter 6
Equalizing Since the hydrocyclones should process enough mud to provide a back flow, the hydrocyclone discharge compartment must be able to equalize to the hydrocyc!one pump suction compaltment. This equalizer opening should be near the bottom of the tanks to prevent sucking the pits dry and for proper hydrocydone operation. The diameter of a round opening (or pipe size) required can be determined by the rule of thumb of dividing the circulating rate by 15, and taking the square root.
Formula 6-1 NOll-Presslfred Flow ThoURh a jJipe
Piper Diameter (inches) =
Max circulation gDln 15
Hydrocyclone Installation Following are some important requirements for the hydrocyclone installation. 1.
Hydrocyclones must be installed in a location where the underflow can be seen easily and checked frequently.
2.
It must be possible to work around the unit. A person should be able to easily adjust, remove, or replace any hydrocyclone.
3.
The feed header must be higher than the highest lllud level in the mud pits.
4.
The solids underflow trough and discharge line must have a downward slope.
5.
The overflow liquid header should end at a point below the top of the pits but just above the lUud level. Ifcarried below the mud level, the overflow header should be vented. This is to let the entrained air break out. It may be necessary to divert the downward flow to prevent it fl·om canying air down to the pump suction.
6.
The overflow header should not be manifolded and in a well designed solids removal system ~- it wi!! not be necessary. It also should not be reduced in size unless the manufacturer permits it, and should not have any unnecessary bends.
6-19
Hydrocyclol1es
Chapter 6 7.
If the overflow header discharge has a deep drop (more than six feet) to the pits, "siphoning" of the hydrocyclones can occur (during operation, not with the pump otl). In this case, solids may noL emerge from the bottom of some of the cones. Chuck l1'ith the mOlIl!/ac/llrer oil/he best way to correct this prohlcm. Install a suction breaker as shown in Figure 6-9.
Figure 6-6 SlIcfiOl1l,,'cl'cens tu Pro/eet J)es;!tef'S
1" pipe handle extending up above mud level ' " or as needed ~ Longer as needed
1/2" or 3/8" Expanded Metal
/
End of Suction Pipe
NOle: A good crew with a good system normally does not need screens.
G-20
HydroGyclones
Chapter 6
Figure 6-7 Hydl'oc)lclol1c j'ced Header Proh/ems
em D~,a.~~_:JII
Good
I~ 3 x HDR ~
Some Cones Spray Discharge Some Cones See No Mud At All Note." Never make a horizontal bend onto a hydrocyclone header. Never place a swedge or reducer less than 3 x pipe diameter in front of a header.
G-21
Hydmcyclol1es
Chapter 6
Figure 6-8 Opel/lull' Header Prob/ems - Slurp Discharge C'allsed by Long S'iphol1 Leg Overflow Header
)
High Vacuum in Overflow Header
Long Siphon Leg
I\~
,If,: l,::, .'
.. ',,1 1\
Slurpi~g
Discharge or No Discharge At All
"
High Air Velocity
Over Header Extends far Below Bottoms of Hydrocyclones
6·22
Hydrocyc/oncs
Chapter 6
Figure 6-9 Ovel.11ol11 Hemier Prohlcl7ls
Nipple DiameteF Header Diameter 4 Breaker
Overflow Header
6" 45"
)
Long Siphon Leg
Solids Free '" ••It,~ \ To Discharge
t~.
nll
,l
11"\'
"1./,: 'I \ .'
Air
6-23
Note: Never Extend Overflow Header Below Mud Level!
Hydrocyclones
Chapter 6
Figure 6-10 fir-'cal' Patterns in
a
b
HydrocJ!c!ones
c
e
d
Top of Solids Bed
Smooth Spiral Grooves
Shoulder from very High Head
Sharper Grooves
Top of Solids Bed No Wear No Wear
Hydrocyclone Wear Patterns In an un-weighted mud with normal head and spray d,ischarge, the overflow parts will out last the underflow parts by 3 to 6 times. The wear encountered will be similar to Figure 6-10 (a). Excessive bead may cause problems with sharper grooving if the cone is in balanced operation. If the head is higher, the balance point moves up inside the cone to a higher point, causing wear there. See Figure 6-10 (b).
6·24
Chapter 6
Hydrocyc/ol1es
Rope underflow has no wear in the solids discharge area but a groove will occur above the non-rotating head. Wear will be rapid in the overflow area because of the abrasive material being returned to the active system. See Figure 6-1 0 (c).
A plugged cone is worse than rope underflow, with the water area moved further up the cone and higher wear in the overflow area. See Figure 6-10 (d). With weighted mud, there is wear at the discharge and overflow area unless the cone is plugged or roping. See Figure 6-1 a (e).
6-25
Chapter 7
Mud Cleallers
Derrick® Mud Cleaners The Derrick Mud Cleaner is a set of hydrocyclones positioned over a shaker so that the underflow from the hydrocyclones is screened and the liquid COlle overflow returns downstream to the mud system. The screen throughput returns to the mud system and the cuttings and barite removed by the screen are discarded (Figure 7-1), A DC1Tick high performance linear motion shaker with 4" hydrocyc1ones (mounted so the cone underflow is processed by the shaker screen) can be used in place of a unitized mud cleaner.
Recommended Use Enough Derrick high performance shakers should be used to screen any significant hole section with 100 mesh screens. If 200 mesh screens can be nlll on hole sections requiring weighted muds, a mud cleaner may not be necessru-y. By placing two, three, or even four Derrick linear motion shakers on the flow line, it is possible to utilize the same screen mesh as would be run on the mud cleaner. DelTick Hi-G Flo-Line® Cleaner 2000's virtually guarantees that all the mud is processed. With mud cleaners, only the hydrocyclones process the full circulating rate and then only if enough cones are working properly. Furthermore. hydrocyclones concentrate barite in the undertlow because of its higher specific gravity. Derrick Hi-G Flo-Line Cleaner 2000's are more dependable to operate and can be utilized a greater portion of the time. Many times, the rental cost for two Derrick Flo-Line Cleaners is approximately the same as the rental cost for a mud cleaner and centrifugal pump package.
What is the Purpose of a Mud Cleaner? The mud cleaner was originally designed to remove solids larger than barite from a weighted mud. The shale shaker removes the largest cuttings. The mud cleaner should remove the remaining drilled solids down to the size of the barite. In ullweighted mud the des iller removes this size range. When the barite is added, and the desilter underflow is screened, most of the barite and some of the drilled solids are returned back to the mud. This allows continual removal of all drilled solids larger then the screen mesh size. The original intent was to start using the desilters when the well spudded. A blank, or solid plate, over the screen area permits all solids to be rejected. When the barite is added to the mud, the blank comes off and the screen goes on.
7- j
Mud Cleaners
Chapter 7
Figure 7-1 Derrick® Mud Cleaner
Other Uses for a Mud Cleaner DelTick mud cleaners have been lLsed successfully on oil muds, both weighted and ullweighted. In weighted and ullweighted applications, the mode of operation is the same as already described. Tn unweighted muds, the oil, or any other expensive liquid phase such as KCl, is conserved. Even though the drilling fluid is not as clean as possible, the economics of preserving liquid that is normally discarded with the desHter underflow may be attractive. The solids separated with the screen have very little associated liquid. These solids can be handled with a front end loader and a plastic lined dump truck. In some areas, ecological considerations suggest using Denick mud cleaners to eliminate or reduce the disposal costs.
7-2
Chapter 7
Mud Clea11ers
How Does a Mud Cleaner Work? When drilling with unweighted muds, the mud ideally passes through a shale shaker screen, a desander bank, and a set of desilters. Particles of decreasing sizes are removed in sequence by these three pieces of equipment. Some drilled solids remain in the mud since they are smaller than the size the hydrocyc1oncs can discard. The DelTick mud cleaner was developed to remove larger drilled solids from a weighted mud. Some larger solids in the mud are normally discarded by the hydrocyclones. This discard in a weighted mud can contain a considerable amount of barite. The fine mesh screen placed under the desilters is designed to return barite to the mud while discarding drilled solids. This is the mud cleaner concept. With experience, other applications have also become obvious. The typical separation curves and typical particle size distributions present in muds are shown in Figure 7 -2.
Drilling tluids can contain water, well dispersed bentonite, barite, chemicals, and very fine drilled solids after passing through all of the solids control equipment. Bentonite pruticles, when properly dispersed, are generally less than one micron in diameter and are returned to the mud by a desilter. While drilled solids generated in a weighted mud may vary somewhat [rom those in unweighted muds, there are frequently many drilled solids in the same size range in both types of muds. Variations in the drilled solids are also due to use of button or diamond bit.s, or because of large pressure differentials. Before the advent of the mud cleaner system, the centrifuge was the only mechanical solids removal system used after a weighted mud passed through the shaker. Desanders and desilters are not normally used to clean drilled solids from weighted muds because the loss of barite in the underflow is not economical. They may be used to remove drilled solids from a very "dirty" weighted mud. I-Iowevcr, a Derrick centrifuge can discard particles smaller than most of the bru'itc. This means that liquid. chemicals and bentonite are discarded along with these solids in the fraction of the fluiu stream rejected by the centrifuge. When drilling highly reactive clays, a centrifuge is essential to keep the concentration of colloidal drilled solids at a manageable level.
7-3
Mud Cleaners
Chapter 7
Figure7-2 Mechanical Solids Control (!f Weighted Drilling Mud Centrifuge
Linear Motion Shaker
Mud Cleaner
Removes lowgravity solids and barite smaller than 3.5 microns
Removes drilled solids large r than 74-105 microns before they degrade into ultra flne solids which cause viscosity build-up, Small amounts of oversize barite may be removed with drilled solids.
Returns larger solids to active mud system. Operates on small fraction of total mud system.
Returns smaller solids, including barHe, to active system. Operates on entire circulating volume.
Removes drilled solids larger than 105 microns before they degrade
into ultra fine solids which cause viscosity build-up. Small amounts of oversize barite and some colloidal barite may be removed with drilled solids.
, REGION WHERE I
Returns smaller solids to active mud system. Screens entire circulating volume.
Screens entire circulating volume.
t
+-
Removes drilled solids large r than 250 microns. Small amounts of barite may be removed with drilled solids.
Returns smaller solids, including barite, to active mud system.
'---~--------'I 3 TO 1> MICRONS, MINUS
Scalping Shaker
.-____ ~+~_____ I
r* 74 TO 105 MICRONS PLUS
~
,
*
25 MICRONS
I DRILLED SOLIDS CAN I
t
1
I NOT BE REMOVED
~ FLC 2000
I Flo-Line
'FLC
r?~'~~~p~<~.~;~:' 325
200
150
100
80
60
7-4
Chapter 7
Mud Cleaners
he mud cleaner system was designed to remove drilled solids larger than barite. The Illud is first passed through a shale shaker. The screens should be as fine as possible and still accommodate the full circulation rate. Then. mud is processed through a desilter bank. The overflow from the desilters is returned to the mud system and the underflow is allowed to fall onto a fine mesh vibrating screen (generally 200 to 230 mesh), Solids larger than most barite are rejected by the screen. Both the majority of the barite and the fluid, as well as drilled solids finer than the screen size, are returned to the mud system. If fine enough screens can be run on the Derrick Hi-G Flo-Line Cleaner 2000's, a mud cleaner,may not he needed. Barite is generally ground to a size range between 2 and 60 microns (Figure 7-3). API Specifications (API Spec 13A) on barite size states 3% by weight is the maximum amount permilted larger than 200 mesh (74 microns) and over 5% by weight must be retained on a 325 mesh (44 microns). Standards in API RP 138 indicate that any material retained on a 200 mesh screen (larger than 74 microns) will be called "sand". So, in every 100 Ibs. of barite, 3 l11s. could be "sand", by API definitions. The problem in any weighted mud is to dean the fluid without removing too much barite. Obviously, most barite will pass through the coarse mesh screens used on scalping shakers. An 80 X 80 mesh screen, for example, has an opening of 177 microns. When the drilling fluid from the well passes over the screen, the large drilled solids (hoth clIttings and cavings) are removed by the shale shaker. Other drilled solids pass through the screen openings (Figure 7 -4). The size range of drilled solids generated in a hole depends on many factors. The bit action and type are the predominant ones, Even with diamond bits, some cuttings can be large. With good hydraulics, adequate weights and rotary speeds, large cuttings m'e often generated. Many of the drilled solids come from the side of the hole. In some holes with long, uncased intervals, the volume of solids entering the mud stream from the side of the hole is as great, or greater, then the solids volume generated by the bit. Some formations, however, contain highly dispersible clays that appear to almost dissolve in the mud, These are usually known as "mud making" shales and the solids generated can only be removed by a centrifuge, After the mud passes through the shale shaker, the mud cleaner hydrocyclones remove most of the larger particles left in the mud. These large particles, barite and drilled solids, fall on the fine mesh screen, Most of the barite and some of the drilled solids pass through the fine mesh screen and return to the mud, The screen rejects drilled solids and also some CalTy-over barite (Figure 7 -5). The carry-over (or piggyback) barite can be significant, and should be monitored closely. Calculation 7 -1 and 7-2 can be used to analyze the economics of the mud cleaner,
7-5
Mud Cleaners
Chapter 7
Figure 7-3 Typical Barite Size Distriblltion for Commercially Available Barile 0 0 ~
x t§2
"on " '" ~ "8u
800 -
600 -
C
--'"
,~
400
~
c0)
u
c
0
U 0)
200
>
'C
'"
03 ~
a
I
o
10
20
30
40
50
60
80
70
90
Barite Size in Microns
Figure 7-4 Sample Distributioll of Drilled Solids ill Mud Ajier Passing Through an 80 Iv/esh Screen
I a
50
100
Mud Passed Through +- 80 Mesh Screen on Shale Shaker
150
200
250
Barite Size in Microns
7-6
Mud Cleaners
Chapter 7
Figure 7-5 Solids Removed by Mud Cleaner
I
.
Badte
a
Mud Through Mud
+- ~~~~~rs~a~~~e:;th
\
200 Mesh Screen
I
+
Romoved with 80 x 80 Mesh Shale Shaker Screen
.~
50
100
150
200
250
Bru'ite Size in Microns Mud cleaners were initially designed to remove drilled solids from a weighted mud. A centrifuge also removes drilled solids from a weighted mud. The difference is the size range of particles removed by each piece of equipment. A mud cleaner removes drilled solids larger than the barite. A centrifuge removes drilled solids smaller than the barite. They complement, rather than replace, each other. Drilled solids appear in all size ranges. After
7-7
Mud Cleaners
Chapter 7
Figure 7-6 Mud 1I'ilh High %
~f
325 200 150
-+-
I
I
Ultra-Fine Solids (<20 Microns) Should be Celltrifuged 80
I
I
DRILLED SOLIDS
Be,,,",
.\ o
50
100
150
200
250
Barite Size in l\1icrons
Figure 7-7 Bentonite Size Distriblltion Superimposed Over the Distribution Presel1ted in 200 Mesh Screen
Barite Size in Microns
Location of Mud Cleaners The Derrick mud cleaners are located all the mud tanks in the same location as the desilters. The hydrocyclones on the mud cleaner can be used during the unweighted phase. A blank plate, or solid sheet, CUll be mounted where the screen is normally mounted. In this manner, all of the solids discharged from the des iller cones are discarded, When barite is added to the mud, the solid sheet, or blank plate, is replaced with a fine mesh screen.
7-8
Mud Cleaners
Chapter 7
The centrifugal pump supplying the Derrick mud cleaner desilter cones should have the suction in the compartment upstream from the dcsiller discharge. Be careful with the rule-of-thumb that says the mud pumped through the desilters must be 10% more than the flow rate pumped down the hole. Any mud pumped into the mud tanks in front of the desilter discharge must be pnssecl through the desilters. Therefore, any mud gun flow rates must also be accounted for. The correct equipment locations arc shown in Figure 7 8. The logic of the equipment layout follows these simple concepts: The mud is treated in an orderly sequence; Each pump performs one task; No two pieces of equipment arc in parallel; and All suction tanks are well agitated.
Figure 7-8 Locatioll of Mud Cleaner Return to Hole
-.<"
Mud Cleaner
'"<@"§!JJ
Screened <.SolIds
7-9
0
---
,( DERRICK ""
,~~.
EQUIPMENT COMPAlIY
Mud Cleaners
Chapter 7
Guidelines for Operating Mud Cleaners L During "weight-up" of a mud system, a large amount of barite may be added at one time. For example, to raise the weight of 2000 bhls from 12.0 ppg to 13.0 ppg would require adding about 127,000 lbs. of barite. If 3% of this is larger than 200 mesh, at least 3800 lbs. could be easily thrown away by the mud cleaners if this fresh barite is passed through. Also, after one or two circulations, this barite reduces to a size which will pass through the mud cleaner screens. For these reasons, some drillers shut off their mud cleaners during weight-up, particularly if they are not drilling. Some barite will be lost by piggybacking on the drilled solids and/or larger barile. If the barite is added downstream from the solids removal section, and not recirculated in the tanks, less barite loss will occur. 2. The most frequent complaint heard by someone using a mud cleaner for the first time is -- that it is throwing away all of the barite. When solids are removed from the mud, the mud weight decreases whether these solids are barite or drilled solids. By measuring the screen discharge parameters, (tlow rate, Illud weight and percent solids), the amount of barite being discarded can be calculated. Calculation 7-1 and 7-2 are useful in determining the economics of operating a mud deaner.
Additional barite is required to keep the mud weight constant. Some barite does cling to the drilled solids, which is what usually accounts for the barite losses.
3. Hydrocyclones are designed to operate at a constant head. If a pressure gauge is installed on the manifold, be sure that the pressure is increased if the mud weight is increased. With electric molors running at a constant speed, the centrifugal pump will do lhis automatically. TIle pressure should be increased by a ratio of the mud weight (MW) of the mud to the weight of water (8.34 ppg). For example, to maintain 75 ft. of head on a bank of hydrocyclones operating on waterrequires a pressure (P) of 32.5 psi to operate properly. If the mud weight is increased to 15 ppg, the pressure on the desilters should be increased to 58 or 59 psi: po::: 75ft x .052 x 15 ppg = 58.5 psi
This pressUl'e will be required to produce the constant head of 75 ft. Further discussion of head and pressure may be found in the section on Centrifugal Pumps. An approximate rule-of-thumb could be developed here for hydrocyclones that require 75ft. head. A pressure of about four times the mud weight will result in slightly over 75 ft. of head at the hydrocyclone manifold.
7-10
Mud Cleaners
Chapter 7
Formula 7-1a cS'!urry-Density, Head-Pressllre Relationship
Pressure, PSI
= Head, ft.
x .433 PSIIft x slurry, ppg/S.34
ot'
Pressure, PSI
= Head,
ft. x .052 x slurry density, ppg
Ca/culatc the head or COl1l1l111 height that would be supported by a /(1/0)1'1/ pressure: Head, ft = Pressure, psi/.433 x 8.34/sJuITY density, ppg Head, ft = Pressure, psi x 19.26/s1urry density, ppg
Slllrry dellsity pressure relationship ill standard imernafioT1allll1ifs. Pressure, kPa = Head, m x slurry, Kg/m-l x .0098 Head, meters = Pressure, kPa x 102/S1UlTY Density, Kg/ml
Formula 7-1b Feel of Head Requiredfor Proper
Hydrocyclone Operation p
~
(75) (.052) (mud weight)
P = 0.9) (mud weight) or approximately P = 4 (mud weight)
7-11
Mud Cleaners
Chapter 7
4. Make certain the mud cleaner desilter cones receive mud from one compartment and discharge mud into a compartment downstream. No other piece of equipment, such as a degasser. should parallellhe mud cleaner. In other words, all mud must enter a downstream compartment by being processed through the mud cleaner. An underflow equalizing line should connect the two compartments and mud should flow from the downstream compartment to the upstream compartment through the equalizing line. The material passing through the screen may be dumped into either compartment as long as it enters where it will be well-mixed. If these solids are dropped into a dead spot in the tank, they will settle and not become a part of the mud system. 5. Most 4-inch desilter cones process between 50 and 60 gpm of mud at 75 ft. of head. A sufficient number of cones should be lIsed to process all of the mud. Be sure to account for any mud being pumped back upstream through the mud gun lines. 6. No solids coming off the screen usually means that the screen is torn the mud is clean.
7. a.
not that
Mud cleaners can be lIsed as a desilters (no screen) until the barite is added to increase the mud weight. When barite is added, install a fine mesh screen (200 to 230 mesh).
b. Mud cleaners may be used on unweighted mud to reduce the amount of liquid discarded by the hydrocycJones. 8. Continuously lUn the mud cleaner. It is designed to treat the whole mud system all of the time. Even in slow drilling, sloughing or cavings from the wall of the borehole increase the undesirable drilled solids loading. If solids are being removed by the primary shale shaker, the mud cleaner will also remove solids. Once the solids have gone back down the hole, the mud cleaner will probably not be able to remove them with a second circulation. 9. If water is prescribed for a heavy weight mud, it can be added effectively over the screen. This will decrease the barite loss from the screen. However, if water is added to the screen, remember barite must be added to increase the density of the water to the required mud weight. 10. If solids collect in large patties, the screen is usually "dead" in that region. This means that it has lost tension and should be corrected.
...
~.
, DERRICK ~ " ,.
._ ,"
EQUIPI.IEIIT COMPANY
7-l2
Mud Cleal1ers
Chapter 7
Calculation 7-1 Low Gravity Solids Concentration in Drilling Mud I. 2. 3. 4.
Retort mud to obtain volume percent solids (Vs ) Weigh mud after air removal. Select barite density and low gravity solids density. Calculale volume percent low gravity solids (V s) . Low gravity solids include drilled solids and bentonite from the equation below.
Barite Density
Low Gravity Density
Qn
Qu-;
Equation
V,,;~58.82+1.96
V,·7 .06 MW
~60.61+2.00
V,-7 .27 MW
4.3 ............ 2.75 ......................
V,,,~64.52+2.13
V,·7 .74 MW
4.2 ............ 2.60 .......................
V,,,~62.50+2.00
V,·7 .50 MW
4.2 ............. 2.65 .......................
V,G~64.52+2.06
V,-7 .74 MW
4.2 ............. 2.70 .......................
V,G~66.67
+2.13 V,·8 .00 MW
4.2 ............. 2.75 ......................
V,,,~68.97
+2.21 V,.8.2 8 MW
VLG
Volume Percent Low Gravity Solids Volume Percent Solids Mud Weighl (PPO) Density of Water Density of Low Gravity Solids Density of Barite
4.3 ............ 2.60 .......................
4.3 ............. 2.65 ...................... V,o 4.3 ............. 2.70 ...... .
V, MW
Qw QLG
QG
These equations result frolll substitution with the formula:
( 100Q,,') + (Q,,' Q,,) V, ( Q,' Q,,,)
7-13
( Q,,' Q",l
( 12 )MW (QB-QLU)
Mud Cleal1ers
Chapter 7
11. Cones continuously plugging usually means that the mud box in the back of the shale shaker is being dumped into the mud tanks before trips, or the shale shaker is being by-passed, or the screens are torn. All mud entering the tanks should go through the shaker screens first. If this is done, no solids in the system should plug the hydrocycloncs. 12. Generally, the mud cleaner will be installed where a desilter bank norm
Calculation 7-2 Drilled Solids (llld Barite Lossfrol71 Mlid Cleaner Screens 1.
Measure time to fill quart mud
2.
Calculate flow rate in gal/min. (4 quarts = I gallon).
3.
Weigh solids with mud balance.
4.
Retort solids to determine volume percent of solids (V s )'
5.
Calculate volume percent low gravity solids (VLG ) from equations in step 4 of Calculation 8-1.
6.
Subtract volume percent low gravity solids (VI.G) frol11 retort solids (V s) to determine percent barite (V n)'
7.
Calculate low gravity (or drilled solids) discard in lbs.lhr. with the formula: 13.8 x (flow rate, gpm) x (V LC). This assumes a 2.6 density for the drilled solids.
8.
Calculate barite loss in l11.1h1'. by the expression: 21.0 x Wow rate, gpm) x (V B)' This assumes a 4.2 density for barite.
CLIp
with solids removed by screen.
Chapter 8
Celltrifugal Pumps
Centrifugal Pumps An important concept about centrifugal pumps is that a centrifugal pump is not a positive displacement pump like a piston pump or progressive cavity pump. A positive displacement pump will pump a certain amount of fluid for each stroke or rotation. A centrifugal pump docs not work like this, Output of a centrifugal pump depends on several factors, The following information and concepts should assist you in understanding how a centrifugal pllmp works.
Slurry Density-Pressure Head-Pressure Relationship If the height (head) of a hydrostatic column is known, the pressure exerted at the bottom of that column can be calculated by the following formula (API units):
Formula 8·1a Pressllre Head Pressure (psi) = Head, ft. x .433 psilft x Silmy Density. ppg 8.34 or Pressure (psi) = Head, ft. x .052 x Slurry Density, ppg
The above is the formula that we use to calculate the bottom hole pressure a drilling fluid exerts on a formation where height is the true vertical depth. For a mud weight of to ppg and a true vertical depth of ! 00 feet the pressure is as follows: PSI = 10 x .052 x 100 or
52 PSI
To calculate the head, or column height, that would be supported by a known pressure we need to reanange the formula as foHows:
8-1
Centrifugal Pumps
Chapter 8
Formula 8-1 b PreSSllre Head & Equivalent Head
Pressure Head, ft. = Pressure, psi x .433
8.34 ShUTY density, ppg
or
Pressure Head, ft = Pressure, psi
x
19.26 SIUlTY Density, ppg
or Pressure Head, ft =
PSI ShllTY Density, ppg x .052
For standard illternaliol1a/ Ul1its, the formula would be as follOll's: Pressure, kPu = Pressure Head, m x Slurry Density,
Kg/nr~
x .0098
When the pressllre is knOll'11, the equil'{l/el1f head is calculated
asfollol!'s: Pressure Head, meters =
Pressure, kPa x 102 Slurry Density, kg/ml
Centrifugal pumps are constant head devices, not constant pressure or constant output devices. A centrifugal pump will output a constant feet of head or pressure head, this is an equivalent height of a column of fluid that it will support. If the slulTY density increases, there will be an increase in pressure at the pump Olltput, but the head stays the same. For a graphic illustration, see Figure 8-1. A centrifugal pump accelerates the fluid presented to it to a velocity near that of the tips of the impeller. This velocity can be varied by changing the RPM of the pump or changing the diameter ofthe impeller. Since the same force of gravity that accelerates a falling object will decelerate a rising object we can calculate the height to which a centrifugal pump will lift fluid (feet of head) using the following formula (neglecting friction);
8-2 EQUIPMENT COMPANY
Chapter 8
Cel1trifugal Pumps
Formula 8-2 Felocity V=(2gH)" Where:
V = vertical velocity g = gravity acceleration H = vertical fall distance
This is a/orm (d'fhe equation: V'=2gH or
If a centrifugal H=
pUlllpS
H=V' 2g
impeller tip speed is 80.2 fps then we have the following:
LIill2l' = 100 feel
or
100 feet of head
2(32.2)
Figure 8-1 How Cenfr(fugal Pllmps Produce Pressure - Relationship Between Vertical Height (Pressure Head) Gnd ImjJcller Tip Ve/odfy
Slurry Density 8.34ppg PSI = 36.9
8-3
Slurry Density 15ppg PSI:;;: 66.3
Chapter 8
Centrifugal Pumps
Figure 8-2 WcJrk
"S.
Shl/t~ill
Friction
--101ft. --99ft
10W
100 FT. SHUT-in
HEAD
AMMETER
v
80.2 fps
Since the velocity of the fluid leaving the impeller tip increases directly with rpm and pressure head varies with the square of velocity of fluid, then pressure head varies directly with the square of rpm.
Formula 8-3 Centrifl/gal Pump Pressure Head HEAD, = (HEAD, x RPM)' (RPM,)
DERRICK fQUIPMEIiT COMPANY
8-4
Chapter 8
Centrifugal Pumps
Work vs. Shut-In Friction Set up a pump with an impeller tip speed of 80.2 fps, and connect a 99 ft. standpipe, with a 2 n. collar. Next introduce water to the pump at 10 gpm (Figure 8-2). The pump will accelerate that 10 gallons of water to 80.2 fps. Every minute the ammeter needle will move from air friction, Fa, to a higher position, indicating a rate of work done accelerating that volume rate of water. The \vatcr will move up the standpipe to a height of:
H
=V' = 2g
80.2' 2(32.2)
100 ft.
The water will reach the calculated height and then stop. It can go 110 higher, the pump can not take any more water, the water in the pump housing will just circulate.
In this situation no work is being done, therefore the ammeter needle will drop to a point which represents water fl;ction. TIle pressure gauge reads the hydrostatic pressure of 100 fL of water or 43.3 psi. The shut-in head of the pump is 100 ft., which could have been determined by closing a valve and pumping against this closed valve. Converting the pressure would give an equivalent head of 100 ft. If this is done again but with a drilling fluid weighing twice that of water (2.0 s.g. or 16.68 ppg), the results will be the same except that the ammeter will indicate twice the work rate, The horsepower requirements for the pump motor will be twice that of what is needed for water, and the pressure gauge will read higher (86.7 psi). If the two (2) foot long collar at the top of the standpipe is removed, leaving only 99 feet of pipe, the 100 foot column of mud will begin to overflow. If mud is now introduced to the pump, it will pump at a rate that will use lip the OIle available foot of head in flow losses. The ammeter indicator will also point to a higher point, indicating that work is being done.
Why is More Horsepower and More Pressure Required for Heavier Fluids? Ittakes more horsepower and fuel to push an 18 wheeltfuck down the road than a small compact car. It also takes more energy and horsepower to transfer heavier fluids because more weight is being moved in a pipeline.
8-5
CCl1triji/gal Pumps
Chapter 8
We are assuming below that a system requires 85 feel of head to move 1000 gpm of water. The S,llne pressure head (85 feet) is required to transfer 1000 GPM of any other fluid. As we have seen, 85 feet of another fluid is not the same pressure (psi). The weight of one gallon afwater is 8.34 ppg. TIle weight of 1000 gallons of water is 8340 lbs. In the example, 16.7/GAL mud is being lIsed as the other fluid because it is 2 times heavier than water and therefore, easier to understand. The weight of 1000 gallons of 16.7 ppg mud is 16,700lbs.
Table 8-1 E);ample of POllnds Tmnsferrallcefor Wafer
I'S.
Mud
Fluid
Pumped
Pounds Transfen'cd
WATER
1000 GPM
8340
MUD 16.7 ppg
1000 GPM
16700
Volume
It takes more energy to move 16700# than to move 8340#. There is more energy available; 85 feet of water is 36.8 psi and 85 feet of 16.7#/gal mud is 73.6 psi. Since twice the pressure is produced by the pump, it must be doing twice the work and need twice the horsepower. Therefore, horsepower required varies directly as the fluid specific gravity. This means you should COll'ect the horsepower for other fluids by multiplying the water horsepower in the pump cllrve by the specific gravity of the fluid to be pumped.
Pump Flow Rate Flow rate is the amount of liquid going through a pipe in a given time. If a hose stream fills a lO-gallon can in 2 minutes, then the flow rate is 5 gpm, or 10/2. The bottom scale on a cenuifugal pump curve is measured in U.S. gpm. (It is often incorrectly labeled capacity in U.S. gpm.)
What Volume Will leave the Pump? If we think of the pump being a water tower filled with fluid it is easier to understand how it works. The pressure head would vary with height of the tower.
The amount of liquid that ran out of the tower would increase lIntilthe losses in the total pipe system equaled the pressure head produced by the waleI' tower. The same is
8-6
Chapter 8
Cel1trifugal Pumps
true of a centrifugal pump. The volume coming out of the pump increases until the system head loss equals the pump output pressure head.
Table 8-2 Example (~f Pump Head Gnd System Loss Pump Head
500 750 1000 1200
System Loss
105' 103' 100'
25' 100'
98'
144'
56'
In the example, the pllmp \l'Oldd operate at 1000 gpm because that is the only point the system loss equals the pUlllp OlllPllt head. The only exception would be when the liquid will not run into the suction as fast as it is being pumped out to the discharge. At that point, the pump would cavitate, loose output head and no fUlther increase in volume would OCCtIl'.
How Do You Know if the Pump is Producing the Correct Pressure Head? Pressure gauges do not measure pressure head, they measure pressure, and the pressure is always changing with mud weight. Most mud systems require 85-90 feel of pressure head at the pump. At the des ander, desilter, mud cleaner, hopper ancl mud guns, you need about 76 feet. Ifwe assume these are correct for your rig, then the pressure reading in psi on a gauge should be 4.4 times the mud weight in ppg at the pump and 4.0 times the mud height in ppg at the equipment. This means the pressure gauges at the pump would read 44 psi and at the equipment, 40 psi, if the mud weight was 10 ppg. With 11 ppg, you need 48.4 psi at the equipment, and 44 psi at the cones.
Total Dynamic Head in a Low Pressure Mud System To calculate the total dynamic head required of a centrifugal pump for proper operation of a piece of equipment (hydrocyc1ones, degasser, etc.), there are several factors Ihat must be taken into consideration.
8-7
Centrifugal Pumps
Chapter 8
As an example, a desilling unit will require a certain feet of head at the manifold for proper hydrocyclone performance. Anything more, or less, will reduce the efficiency of the unit. If a desilting lInit requires 75 feet of head, a centIifugal pump mllst compensate for the ftiction losses in the lines and fittings as well as for the lift or fall from the mud level in the pits. A pressure gauge at the manifold will show the psi supplied and the feet of head can be calculated, provided you know the mud weight. If it is low, then a larger impeller is needed, or perhaps the plumbing could be modified to reduce the friction losses. The follo\ving drawing shows a simplified system setup.
Figure 8-3 Toted Dynamic Hem/for Centr(fuga/ Plimp Feed Head
Friction
> Formula 8-4 Total Dynamic Head
Feed Head + Lift (or-Fall) + Friction = Total Dynamic Head
8-8
CCl1Irifugal PUII1PS
Chapter 8
Cavitation In a centrifugal pump, the eye or inlet is the point of lowest pressure. TIle highest pressure is in the casing just a few inches away from the inlet. Cavitation begins with the formation of vapor bubbles in the eye during pumping. Several factors can cause vapor bubbles to form:
• • • •
The pump's suction
~ine
is restricted;
The t10w rate through the pump is too high for the inlet size; TIle level of mud in the pit is low compared to the sllction line; and The mud is very warm.
Figure 8-4 Cel1trifilgal PumjJ Cavitation
n. b. c. d. e. f.
Eye. or Inlet Impeller Casing Discharge Vapor Bubble, Formed Vapor Bubble, Imploding
e
f c
8-9
/
Centrifugal Pumps
Chapter 8
The lower pressure or vacuum will cause the mud to boil at a temperature much lower tllUl1 usual. When cavitation occurs you will hear a sound similar to gravel being pumped through the pump. TIle noise is caused by the vapor bubbles collapsing or imploding (opposite of exploding) as they pass rapidly from the center eyelet of the pump to the higher pressure area between the casing and the impeller. At the point of implosion, a pressure wave is created that is very destructive to any metal surface close to it. The impeller, face plate, and pump hOLising can be seriously damaged in a matter of minutes. If you ever disassemble a centrifugal pump and the inside looks like Swiss cheese, you know that the pump was cavitating. If a noise indicating cavitation is heard, shut down the pump immediately and check for damage.
Figure 8-5 Rig Up Problem - Close Suction Ell
SUCTION FLANGE
IMPELLER VANE
GOOD~_r-
'DERRICK ~,-
- EQUlPME", CO/,lPANY
8-10
Centrifugal
Chapter 8
PUI11PS
Rig Up Problem - Close Suction Ell (Elbow) When a liquid flows through an elbow, the liquid is thrown to the outside of the pipe due to centripetal force. Several diameters of pipe are required after the ell for the flow to regain a uniform pattern. If the flow enters a pump directly from an elbow, it throws an unbalanced load on the impeller and tends to cause cavitation on the inside of the discharge and erosion on the outside. This unbalanced condition causes severe vibration and fatigue on the shaft, and leads to packing and bearing problems. See Figure 8-5 for correct placement of elbow.
Line Sizing and Line Scour Velocity When drilling mud flows in a horizontal or upward slanted pipe, solids are continually settling to the bottom of the pipe. If the velocity is high enough, the solids will be scoured along the pipe until reaching a vertical section. In a vertical section. the solids can be transported at a much lower velocity (70 ft/Olin or 1.2 fps).
Figure 8-6 Seltlillg of Solids ill Pipe
o
o A. Original Problem
B. Solution #1 Unsuccessful
c.
D.
Solution #2 Still Worse
True Solution
When the velocity in a horizontal pipe drops below the scour velocity, the solids will settle to the bottom in a stahle position and stop. More solids will settle on top and form a very stable bed in the bottom of the pipe. This bed is very difficult to remove and cannot be removed by the available pressure in the rig's low pressure system (Figure 8-6).
8-11
Centrifugal Pumps
Chapter 8
The hard bed forms under dynamic conditions when the average cross-sectional velocily of the mud drops below 300 ftlmin, or 5 fps (1.5 m/s). Velocities above this will prevent settling.
Figure 8-7 Hard Bed Formation - Velocities Above 300ff./min. Will Prevent Settling
To calculate the cross-sectional velocity of nuid in a pipe use the following fonnula:
Formula 8-5 Cross Sectional Velocity of Fluid ill ({ Pipe Velocity, fps =
now rate, GPM x (.4087) pipe diameter 2
On suction lines, too high a velocity should be avoided because this will cause cavitation. Velocities above 9 fps on suction lines should be avoided. On pump discharge lines, the same rule applies for scour velocity, to avoid settling in the pipe. High velocities on the dischmge side will not cause any problem with the pumps, but there can be excess wear especially in 90 degree ells, branched tees, etc. The upper limit on the discharge lines should be 12 fps. The most common cause of cavitation is slowing of the pumping rate, resulting in formation of a hard bed that reduces the suction line cross-sectional area to a diameter too small to maintain the flow rate. When the pump speeds up to the higher rate, the restricted suction pipe causes cavitation. To avoid cavitation, it is recommended that one pump be sized for each specific job. Manifolding pumps together so that any pump can run any piece of equipment is asking for trouble.
Chapter 8
CClllrifugal Pumps
Figure 8-8a Friction Loss ill Feet (~f Head 10,-------r---------------~------_.
CIilB4"L-4"T
- - - - - - - -
S"L_SUT
-,-
- - --
"
~OL----~----~---~-----.J 200
300
400
500
600
GPM
Figure 8-8b Friction Loss in Feet of Head 5
------,------
-----,--------~--------
o 800
1000
1200
1300
GPM
8-13
1500
2000
Chapter 8
Cel1trifugal Pumps
Figure 8-8c Friction Loss ill Feel of Head
7 _______ J ________________ 1 ______________ _
,
,
8"
_____ 1. _______ -'
6" L
__ 8"_L __
-------4--------1--------~-------4--------
,
"
o 500
600
800
"1000
1200
'1500
GPM
Line Sizing & Friction Loss For some centrifugal pump applications, the line size may not be critical. but for proper hyurocyc1one operation (desanders and desilters), the line sizing is critical. On low pressure systems, we recommend that one pump operate one unit, and if this
is the case, the size of the suction line should be the same as the discharge line. Unfortunately, this isn't the case on many rig installations. One of the most common misconceptions seems to be that if you have an 8" x 6" centrifugal pump, the suction line is 8" ami discharge piping up to the hydrocyclone unit is 6". If a 8" pipe is needed for the suction. then you should swage up to 8" at the pump discharge flange and continue the piping with 8" line to the hydrocyclone manifold. Another problem is that often one centrifugal pump is sel up to run one or more pieces of equipment. Proper line sizing (as well as impeller sizing) for this is impossible. The pump line sizing calculator assumes one pump for one hydrocycJone unit. The calculator pickS the proper line size, without allowing input of improper line sizes, You can calculate this manually if need be,
Chapter 8
Cel1frijugai Pum/,s
The velocity in the pipe should be 5 fps or higher to eliminate settling and forming a hard bed in the line, and 9 fps or less to minimize friction losses and line wear.
Use the preceding graphs to decide \vhich is the proper pipe size for your flow rate (gpm).
Friction Loss in Fittings (in equivalent length of straight pipe) The data in the graphs above are ca1culateclusing data from the National Hydraulics Institute. The data shows the friction losses for equivalent straight lengths of pipe of the given diameter. In calculating your friction losses, inventory the fixtures and fittings in your system. Pipe finings are couplings, elbows, lees, reducers, valves, and the like. They are made threaded, welded. and flanged. When the fluid flows through the fitling, the velocity pattern is disturbed, and this causes a friction loss of an amount high enough to include in pipe tliction calculations. The greater the velocity disturbance, the greater the friction loss. Welded fillings me smooth on the inside and have less friction loss than threaded fittings. Long-radius elbows have less friction loss than standmd or sqllare elbows. TIle flow is smoother in long turms.
8-15
r~1 ~! '::0'
C') :T
~::o ~-
2"
~(")
~"
g ~
~
-
Table 8-3 Friction Loss in Fittings - API Units In Approximate Equivalent Feet of Straight Pipe
9QoWelding Ell
=
6:12
2:3
Submerged Entrance
L = Laminar Flow T = Turbulent Flow
10"
H"
T
q
T
LI
T
L
T
17
23
32
6
4%
7
"
25
3V,
29
16';; 2 2Y, 2Y,
18 3
3 10 43 27
29
4 3 2 2
19 5
45 8\'; 6;" 12 8 4 36 5 5~ 49 3\1, 31 2',"- 22';, 3 3 3% 3;1, 4;1, 4;1, 5
4/, 4,-;
6 3'1,
7
7'1;,
71,
9
,
10
10 11\1 11;1, 12'10 lZY; 4';'
5~,-
5',~
6
6
4
4
4'/, 4'/0
2'1,
2'/,
3
I
I
I
5
5
3
3
1;1,
IV,
5
2
'/, NEGL %
~~
NEGL I
I
5:3 - 10:6
4:2Y, 8:5 3:2 - 6:4 .- 12:8
T
L
T
]Y:
4;6 - 8:12
en
5'
23
3:5 6:10 2'/;,:4 -- 5:8 3:4 6:8 2:2V,-4:5-8:10 Reduce** Down Ratios: 4:2"" 6:3 = 8:4 = 10:5 = 12:6
1'''
4'
IS
L
"
45" Welding Ell
3~,-"
L 10 16 12 20 3/'; 2;1, 4 5';' 3 15 2 18 zv, 10 1'/, 11'.1, IV, 14V, 11', II', 2 2 2
T
13
Weld Tee, Running
Gale, or Butterfly Valve Swage"· Up, Ratios: 2:4 = 3:6=4:8 = 5:]0
3' T
L
Weld Tee, Branched·
2~/,"
;::
7
2 y,
NEGl
NEGl I
I
Yo
'I,
I
I
NEGL NF-UL 3!/, 3%
'h
NEGL 4;1,
4'1~
I
5'!.
5;1,
I ;';
y,
;1,
m
,
y,
IV,
5
7
7
L 64
T
55
10 7'1,
17 57
4
35
5
43
5
5~';
5~t,
7
13 9 6 7
T 34
L SO
9
J3
6~',
9
"
48
L 77 21 80 51
T
8'h.
~y,
58 15 II
7
8
I~
~
Q,
O· ;::
l'
0
en en
3
4
4
I
I
I
1,~
y;
'is
'~
2
2
ZV:,
2V,
I
I
'"
18
lV,
IV, I
II
12
12
I
I
y,
'!.
J4
J4
;::
~
8
6
I
II
14;r, 16'/, 16Y,
5
NEGL 9
T
(1)
;:;,
."
~.
y,
NEGL
6
LI
12"
Q)
6
5
6. ;:: (Jq
en
3
3
'!.
y,
2J
2J
Note: Add one-third/or eccentric redllcerS and swages. For swages and redllcers. lise the line size. if both ends are connected /0 lines, lise the .Imalliine size in the lobiI'. Adapledfrom injarmalion in "Cameron Hydraulics Data ", TUbe Turns "Flow oj Fluids ", and Crane Company'; "Hydrarl/ics Book. "
o? ~~ ~
-
'" ~
"
S
0<;
"'-
;;>'
or
'"
~
'"
.,uo l
'F ~
Table 8-3, Continued
S' C/O
Friction Loss in Fittings - API Units In Approximate Equivalent Feet ofStraighr Pipe
~
Q)
9.
:!1
~.
TEES
o·
**Swage
*Running ()
I~ \ !!l i
*Branched
OIl!
110--'6
O:;:fJi
J
II ---;J
(rr-)- - - - l
'---...
_fI1
o :=:r '--->d
:::
b
-.
C/O C/O
c::=D
'" :J1
O. ~ C/O
;j 0-
**Reducer
F ~
' ..l
(j
'?
..,00;
...
S
~O
"-"1l
I
S
~::tJ
~~
c;o
§
'"'
Chapter 8
Centrifugal Pumps
Total Dynamic Head and Pump Selection (API) (See Worksheet on Following Page) 1.
List the feed rate in GPM for pump suction and dischm'ge.
2.
Use the graphs to determine the proper line size for the feed rate. In a good set up, it should be the same for suction and discharge.
3.
Mud velocity is calculated by the fOllowing formula (See Page 8-11, FOllllUla 8-5): Velocity, fps;:;: flow rate. GPM x (,4087) pipe diameter 2
The
Iyillfe
should be betll'een 5 - 9fpsfo/" a proper instal/ation.
4.
Inventory the exact amount and type of fittings, valves, length of straight pipe, elc. and fill in the table, for both laminar and turbulenl flow, to obtain the "total equivalent length of pipe".
5.
Sum up the laminar and turbulent columns for totallellgth of each column.
6.
From the line sizing and friction loss graphs, find the laminar and hlrbulcnt values per 100 feel and fill in the table.
7.
For the total friction loss in each line for each flow, multiply line 6 by line 7 and divide by 100.
8.
From line 7, write down the largest suction and largest discharge loss and add them together to get the total fIictioll loss from sllction entrance to equipment.
9.
Measure the height above (+) or below (-) the normal mud level in the tank to the equipment feed header. This is the head loss or gain due to the vertical lift or fall.
10. The required total dynamic head (TDH) is the equipment feed head desired + the friction loss + the lift or fall. I L From the pump curves (water tables), choose a pump and impeller combination that will give you the necessary TDH to operate the piece of equipment. 12. For the maximum horse power requirements, divide the maximum mud weight by 8.34 to get the specific gravity of the mud and multiply by the horse power required for water. This calculation is for new cones, add 10% for worn cones.
8·18
Chapter 8
Cel1trifugal Pumps
Table 8-4 -- Worksheet Dynamic Head and Pump Selection, API Units Use with Mlid Line Sidng Curves
Pump Suctioll 2. Line size, from curves. inches 3. Mud velocity. ftlsee minimulll 4. To Obtain Total Feet:
8-19
,.
Pump Discharge
Chapter 8
Cellfrifugal Pumps
Worksheet - Page 2 8.
From step 7, largest suction loss plus the largest discharge Joss (or Ft. + ___ Ft.) = ___ Ft. total fIictionloss, sLiction entrance to equipment.
9.
Height from normal mud level in tanks lip, plus, (or down, minus), to the equipment Ft. This is the head loss, (or gain) due to the vertical feed header or inlet = lift (or fall).
10. Required Total Dynamic Head increases through the pump = Equipment Feed Head desired ( __ft.) + Friction ( __ ft.) + Lift/Fall ( __ ft.):::; _ _ fl. T.D.H 11. From pump curves, choose one with a speed and impeller size to furnish the needed Total Dynamic Head at minimum gplll. Impellers can be ordered to the closest 1/8" inch. 12. For horsepower input, use maximum feed rate, chosen speed and impeller size, and either: a) read horsepower from a special mud density curve for that pump, or b) read water horsepower and multiply by the specific gravity (mud-to water density ratio) of the heaviest mud to be used with this installation,
Pump Sizing Notes:
Input H.P. at Mud WI., PPG MAKE
MODEL
RPM
IMP.DIA. T.D.H.
8.34
11.0
18.0
8-20 EGUIPMENr COMPANY
Centrijilgal Pumps
Chapter 8
Centrifugal Pump Installation Guidelines If the hydrocyclones (and other removal apparatus) are to reduce costs, the following guidelines must be observed regarding the centrifugal pumps and pi ping. L Each piece of removal equipment should have its own centrifugal pump. 2. No mud should enter any suction compartment in the removal system except: a. directly from the removal equipment immediately upstream, or b. from an equalizing backflow. 3. TIle pumps, impellers, motor speed, and horsepower must be properly sized and matched to the hydrocyclones for proper feed head. An impeller size "approximately COlTece or "in the ball park" may not work. Read the section on centrifugal pumps carefully and follow the advice of your hydrocyclone manufacturer. 4. Be sure your line sizes are conecL Oversized lines can cause as many problems as undersized lines. 5. When requirements 3 and 4 above are met, your pump will be COlTect for one purpose, not for several. Use it only for that purpose. ,
6. Eliminate aU manifolding possible. Manifolds calise pump trouble and can result in very poor solids removal because the wrong valves are open and closed or washed oUL Read and study the section on mud system arrangements. 7. Eliminate all unnecessary bends, tees, and other fittings frOl11lhe lines. One sllction valve to penl1it pump repair is necessary. One valve on the discharge for starting the pump with discharge closed is advisable. Open the dischmge slowly after the pump is started. 8. If a screen is installed on the suction, be sure to have adequate screen area. A workable rule of thumb is 15 times the pipe diameter: The screel1 should be IJlOllJ1ted so it c{ln he removcd, clc{lned and reinst{llled by olle 111{11111'hile the crew makes a drill pipe cOI1I1('ctiol1.
9. AITangement of the pump suction and equipment overtlow discharge locations must follow the centrifugal pump manufacturer's suggestions to avoid air lock problems. and also should avoid pumping air to the main rig pump suctions.
8-21
Chapter 8
Centrifugal Pumps
Net Positive Suction Head (NPSH)-Required A pump cavitates when there is not enough NPSH-available. How much is needed to avoid cavitation? Careful tests on pump test stands (by closing a sliction valve slowly until enough vapor forms in the impeller to make noise and reduce lhe total head by about 2%) give the least NPSH-available that the pump can use. The reading is called NPSH-required. It is a fixed amount for any particular pump, impeller diameter, rpm, and flow rate. It varies with the rpm, flow rate, and impeller diameter. NPSH-required lines are plotted on pump curves.
Operating Guidelines for Centrifugal Pumps 1. 2. 3.
4. 5. 6. 7.
8. 9.
10. 11. 12. 13.
Centrifugal pumps are constant head pumps - not like constant-volume piston and plunger pumps. The centrifugal pump head remains the same, although the mud weight changes a lot. To make a good selection of centrifugal pump size, rpm, impeller, diameter, and motor size; one mllst choose the equipment La be connected, design the discharge piping, and figure the total head needed at the flow rate wanted. Always size centrifugal pump piping by flow rate requirements of the equipment, not by the size of some flange on a piece of equipment or on the pump. Always choose centrifugal pump-rpm-impeller diameter combinations by the job to be done, not by the pipe size required, nor by "pump capacity". When pumping mud, choose lower pump speeds to reduce the rate of wem' in the casing and impeller (choose larger impeller diameter to make the head). The person choosing centrifugal pump size-q)m impeller-diameter combinations must maintain a thorough understanding of the interaction between those elements. A flooded suction is a great help in priming the pump and preventing cavitation. Use a single suction for each pump handling mud. Do not manifold two pumps together on the sLiction (especially) or on the discharge. This can easily cause serious, expensive problems. Never have two or more centrifugal pumps pumping llU'ough the same pipe at the same time. Never permit air or gas to enter the suction of a centrifugal pump designed to handle dlilling fluids. Use a suction strainer to keep rags, rubber chunks, sticks, wrenches, etc. out oflhe pump impeller. Clean the strainer to avoid clogging. Never change to another diameter of impeller unless someone who is thoroughly familiar with centrifugal pumps has checked the problem and recommended it.
8-22
Chapter 8
Centrifugal Pumps
Never cut the diameter of an impeller down in the field, it will not be balanced. 14. Never replace an old impeller with a new one,jllst because a change in discharge is shown on the pressure guage. If the read (not the pressure) is too low, then it is time to change the impeller. and possibly the casing. 15. Do not ever adjust the suction valve to regulate flow rate or pressure of a pump. The pump will cavitate and destroy itself. 16. Do not open bypass valves in the discharge pipe system to try to reduce the flow rate in any piece of operating equipment. It wastes power and may overload the motor. Removing some cones does not hurl the pump. 17. Never walk away from a discharge valve that you have closed. The water in the pump win boil after it lUns that way for a few minutes. Close the discharge valve most of the way before starting a pump on an empty system. TIlis prevents water hammer (in the hydrocyclones). Always open it fully before walking away. 18. Be sure that the motor is large enough for maximum tlow rate with the heaviest mud that will be used in the pump. 19. A properly tightened packing must leak 2-4 drops per minute. Stopping the leakage entirely will overheat the packing and the water shaft. Repack a pump properly when needed using the manufacturers instructions. (Some mechanical seals with both seal faces made of very hard metal have a long leak-free life on pumps with very stiff shafts.) 20. Grease (or oil) the bearings according to the pump manufacturer's instructions. Re-grease with clean grease regularly (2-4 times each year).
8-23
Cel1trifuge
Chapter 9
Principles Of The Decanting Centrifuge A centrifuge is a device that crcates a centrifugal force, and imparts that centrifugal force to its contents by rotation. The centrifugal force created by the centrifuge is dependent on the diameter and RPM of the bowl. To calculate the g forces, use the following formula:
Formula 9-1 "g" Forcefor Centrifllge
g's = (RPM 2 )(.0000142)(Diameter of bowl, inches)
Example:
24" bowl, 1530 rpm
g's = (1530),(.0000142)(24) g's = 798 gs The performance of a centrifuge fOf a given mud is dependent on the following factors:
1. The g's exerted on the fluid; 2. The retention time in the centrifuge (the longer the mud is in the centrifuge the smaller the particle that can be separated); and 3. Conveyor differential (the faster the differential, thewctter the solids and the more solids are dischru·ged. AIl of the above are factors which are considered in different designs. Some designs perform better for different applications. Designs often vary in bowl diameter, length, RPM operating range, gem ratios, pool depth, etc. The items which can typically be manipulated in a centrifuge operation are the feed rate and dilution (if any), the gear ratio (by changing gear box), the pool depth (hy changing the weir settings), and the RPM (by changing the sheaves), which changes the g force.
9-1
Cel1trifuge
Chapter 9 To calculate the RPM of a centrifuge. use the following method:
Formula 9-2 RPM (l a Centrifuge RPM
size of motor's sleeve x RPM of motor size of rotating assembly
Example:
motor sheave:;;;: 9.5 inches RA sheave = 5.75 inches motor RPM = 1750
Answer:
9.50 x 1750 = 2891 RPM of machine 5.75
Since the RPM determines the size of belt needed, a change in RPM calls for a change in belt size. To calculate the size of belt needed, lise the following formula:
Formula 9-3 Belt Si;.e Neededfor Centr(fuge Belt size = sum of: 1. 2. 3.
Diameterofsheavc"A"x3.14/2=X Diameter of sheave "B" x 3.14/2 = Y Center of RA to center of motor X 2 = W
x + Y + '\\' = Belt length Example:
9.5" motor sheave. 5.75 RA sheave, center to sheave: 32" 9.5 x 3.141 2 5.6 x 3.14 !2 32 x 2
Answer:
14.92 8.79 64.00 87.71
Belt size is 88" or whichever is closest.
M<"Y'"'"'"'
If··· DERRICK ~~/
EQUIPMENT
COMPAN~
9-2
Chapter 9
Centrifuge
Note: Center of motor should be at middle of the slide (adjustment track), If a gear box is changed, remember that the torque setting also changes. Torque seuings for the DE-WOO electric drive are as follows: 125: 1 = 40 1bs 52:1 = 67 1bs
Set feed to 10% below maximum to avoid overloading the centrifuge. To have a centrifuge perform well for your application, you may need to make changes in one or more of the above settings. As the application changes, adjustments must be made for optimum performance. Some of these changes are time consuming and difficult to make, especially at the Jig site. This being the case, many centrifuges are not run at optimum efficiency. A fully hydraulic drive centrifuge such as the
Den'iek DE-WOO FHD allows the above changes, other than the pool depth, to be made quickly and easily. The DelTick centrifuge offers the ability to change operating parameters quickly to adjust for optimum efficiency as drilling conditions change.
Stoke's law A centrifuge operates on the principle of accelerated settling. Particles will settle in a viscous media according to Stoke's Law, which in simple form is written as follows:
Formula 9-4 Stokes Law
11'I1e/'e:
V, = aD'( p, - p,) 116 [I
X
10-6
a = bowl acceleration, in.lsec 2 = 0.0054812 x bowl diameter x rpm2 Vj = tenninal, or settling, velocity p! = feed sluny density, gral11s/cm-' D = pm1icle diameter, micron Jl = feed slurry viscosity, centipoise Ps = solid (particle) density, grams/cm' (grams/lOO cm. sec.) Stoke's Lall' .shoH'S tilar as fluid efficiency decreases.
viscosit~'
ond density
il1crea.~es,
sepa/"{/fioll
If a sphere senles in a liquid of known density and viscosity, and the settling rate is measured. then by Stokes law the density of the ball or sphere can be calculated if the diameter is known. Or, the diameter can be calculated assuming the density is known.
9-3
Cel1lrifuge
Chapter 9
Figure 9-1 Settling
(~f Spheres
in Fluids, Stoke's Lenv
Since in the real world, we don't drill spherical balls, we use the C(luivalent spherical diameter of the i11'egularly shaped particles, which enables us to use Stoke's Law. When the settling rates of ilTcgularly shaped particles are measured in a liquid with known density and viscosity, the E.S.D. can be calculated, assuming or knowing the specific gravity of the particles. Two particles of the saIlle E.S.D., but different specific gravities, will settle at different rates. The heavier particle will settle faster. Two pmticles of different specific gravities will settle at the same rate if their E.S.D. difference compensates. A barite particle will settle at the same rate as a low specific gravity particle if its E.S.D. is I Y2 times that of the barite particle.
Figure 9-2 Settling of Irregularly Shaped Particles in Fluids, Stokes Law
9-4
Centrifuge
Chapter 9
With this in mind, it is important to screen to 200 mesh (74 microns dSO) with your shakers, or mud cleaner, prior to sending mud to a centrifuge with a weighted mud. This will minimize the drilled solids in the discharge which is returned to the active mud system,
Centrifuge Operation Figure 9-3 Cemr(fllge Operation
Liquids Discharge
Solids Discharge
The mud is pumped through the inlet to the feed chamber and slung from the feed ports into the pool. It is accelerated again with a high shearing effect due to the rotation of the bowl. The level of the pool is determined by adjustment of the weirs at the large end of the bowl (effluent discharge).
As the slun'Y flows through the pool in the channels between the conveyor flights, the particles settle according to Stoke's Law, but at an accelerated rate due to the higher G - forces. Sand panicles settle almost instantly, then the finer particles settle. The particles that cannot be settled under the existing settings of the centrifuge wil1 discharge out the overflow with the liquid phase (effluent). The screw conveyor rotates in the same direction as the bowl, but at a slightly slower fate. This difference is determined by the gear box. Typical gear boxes are 80: 1, 125: I, 40: 1, and 52: 1. 125: 1 gear box means that for every 125 revolutions of the bowl, the conveyor turns one (I) less or 124. With this ratio, a centrifuge rotating at 1800 RPM has a conveying speed of 1800/125::: 14.4 RPM.
9-5
Centrifuge
Chapter 9
The settled solids form a cake on the side of the bowl and are conveyed along the outer edge of the pool towards the solids end (small end) of the rotating assembly and are discharged out the solids discharge ports at high velocity. As the solids travel across the beach, they lose their free liquid film due to centrifugal squeezing and drainage.
They are discharged with only the adsorbed moisture, provided the centrifuge is operated properly. This discharge cake is very heavy, and sticky. It will not flow because of the lack of free liquid. It is important to realize this when installing a centrifuge. If possible, install over the place where the solids are to be discharged. If this is not possible and some sort of slide or chute is needed, remember to have at least a 45° down angle on this slide so that it will self clean. If this is not possible, then some sort of mud wash or water wash will be needed to keep the solids from building up and packing off the centrifuge. Another option is a small amount of remix waler added within the bowl housing directed at the solids discharge to soften the sltm-y. Remember to have the solids directed towards an agitator to keep the solids from simply dropping to the bottom of the tank.
Conical and Contour or Cylindrical Centrifuge Bowl Figure 9-4 Centrifuge BOlFll)'IJes (t.
Conical Cel1tr(fllge BOll'l ........-28"
, \
\
j j
...
h. Colllollr or Cylindrical Bowl ""'-28"
'\
/
~' s;~:,1:j::::::(~,~~~~
A conical bowl centrifuge is shaped just as it sounds. The interior of the bowl is a cone. These were the original centrifuges used in the oilfield and some are still used today. (Figure 9~4[aJ). An advancement in centrifuge bowl design was the contour or cylindrical design (Figure 9A [bJ). In this design, the large diameter of the bowl is kept cylindrical for a portion of the total length, then transitions to the normal angle of cone taper for the remainder of the bowl.
9-6
Chapter 9
Centrifuge
When comparing conical and contour bowls of the same length, several things happen. If the pool is adjusted so that the beach is the same on both machines, the pool volume and therefore retention time is greater on a contour bowl. The result is that a contour bowl can handle a Illuchhigher feed rate while still making the same median clItpoint, as compared to a conical bowl centrifuge of the same size.
Applications For drilling fluid processing there are two uses of centrifuges. One is to remove solids from the mud while returning the liquid phase to the active system, such as when llsed with an unweighted dlilling fluid. The other is when solids are returned to the active system and the liquid phase is discarded. For un weighted drilling tluids. a centrifuge is often llsed on the active mud system to remove as many drilled solids as possible. which helps to control the weight or density of the mud. The effluent (liquid phase) containing conoidal solids, such as bentonite and chemical additives, is returned to the active mud system. If it is an expensive drilling fluid, often the hydrocyclone lInders can be processed with a centrifuge to recover as much of the associated liquid phase as possible while discharging the solids. The other way a centrifuge is typically used is to cont,rol the rheological properties of a weighted drilling fluid. This is often referred to as "barite recovery". When set lip in this mode, the centrifuge returns most of the solids to the active system. Assuming the upstream solids control equipment has removed the sand size particles. then the centrifuge will remove or separate a portion of the sill size particles from the drilling fluid. These silt size particles are barite, perhaps some commercially added products, and silt size drilled solids. A centrifuge willnol selectively remove only the barite while leaving the other solids in the mud. The colloidal size particles (2 microns or smaller) such as clays, fine degraded barite, and drilled particles have a very large surface area verses the volume of solids. As the density of the drilling fluid is increased, there is less and less free liquid. When rheological problem.s start to occur (high viscosity, high gels, etc.) there is not enough free liquid in the drilling fluid and you have a solids problem. In a situation like this, the colloidal content of the drilling fluid needs to be reduced. This can be done with straight dilution. but it is more economical to use a centrifuge. The centrifuge will return the silt sized solids to the active system and the effluent (liquid phase) will be discharged along with the colloidal size solids it contains. The discharge is replaced with clean liquid phase. Continue to run the centrifuge until the rheological properties are acceptable.
9-7
Chapter 9
Ce11lrifuge
Other Uses of Centrifuge With Respect to Drilling Fluids 1. 2. 3. 4.
Process screen unders from a Hi "G" Dryer. Process hydrocyclone unders (expensive water-based. or oil-based muds). Reduce the weight of a drilling tluid by stripping solids out. Increasing the weight of a drilling fluid slurry (reducing reserve volume, or weighting up a slug). 5. In oil based muds, two centrifuges can be used in a series - the first to "recover barite", and the second (nllming at higher g's) to process the effluent of the first centrifuge, removing fine solids which cause viscosity and other problems while returning the liquid phases. 6. Clarification, such as with a flocculation system.
Operation and Maintenance of a Centrifuge 1. A centrifuge that is capable of handling 10% of the How rate is usually sufficient for drilling nuid purposes. 2. For optimum separation efficiency, a Marsh funnel viscosity of 35 - 37 sec/qt is desirable. Use dilution water at the feed inlet if necessary. This also holds true for rotary mud separators. 3. If solids are being returned to the system, (barite recovery mode) a small amount of remix water added within the housing at the solids discharge point will soften the slurry and aid in mixing of the solids back into the active system. 4. Always maintain the centrifuge in accordance with the manufacturers recommendations.
Wear Characteristics of Centrifuges External Bowl and Bowl Heads Excessive wear on the solid end of the bowl and bowl head is generally caused by two situations: a)
The solids built up in the solids discharge end of the case cause solids to pack around the bowl and create unnecessary wear. Solution: Dilute the solids in the case, open the case regularly and clean, or eliminate the use of slides when possible. Finally, adjust the "wetness" of the solids.
h)
Not monitoring the wear of the centrifuge consumables. If they wear beyond the safe point, the solids will eventually damage the solid bowl head and/or the solid end of the bowl. These types of repairs can be expensive and cause seriolls "down time".
9-8
Chapter 9
CemriJuge
Liquid Bowl Head There is seldom any wear that is seen on the liquid end. From time to time, the liquid discharge openings may show wear. Fine solids that are not removed, which exit with the effluent, will create this wear. Monitoring these during the regular service of the unit can prevent serious problems.
Conveyor The conveyor has many areas to be concerned with. The majority ofthe wear is in the feed zone area. This is where all of the feed slurry enters the machine. About one or two flights behind the feed ports, through the discharge end of the bowl, is the area where the most tip-wear occurs. This is normal wear, and is based on how abrasive the product is, the feed rate, percentage of solids and the type of wear protection on the conveyor blades. There can be a large amount of wear toward the liquid end of the conveyor. often seen in our smaller centrifuges. This is caused from "over feeding" the centrifuge. The slurry enters the machine in the feed zone area and pushes its way toward the liquid end of the machine because it cannot be discharged fast enough from the discharge end. Sometimes the back of the blades will show wear. The conveyor hub can also have excessive wear. Again. tllis can be caused by overfeeding of the centrifuge, which is preventable by monitoring the feed ports. This can only be done when the unit is torn apart at the time of repair. It is suggested that the length of time the machine is in the field be recorded, as well as the type of product being fed. If highly abrasive material is being run through the machine, the length of time in the field should be reduced. Monitoring the conveyor wear is extremely difficult because the conveyor cannot be seen inside the bowl. Some centrifuges have a wear gauge which if used regularly, and records are kept, you can help monitor the wear of the conveyor.
Internal Bowl TIle traction strips in the bowl can wear out. A symptom of this problem is the centlifuge periodically kicking out. or spitting solids. High torque and plugging can be caused by loss of traction strips.
Problems Can Be Reduced By: Monitoring wear on the consumable parts; Doing periodic maintenance on the case; and Insuring that proper feed and pool settings are used.
9-9
Hi HG" DI)'er
Chapter 10
Using Hi "G"TM Dryers to Reduce Drilling Costs Mark C. Morgan Derrick Equipment Company, HOllston, Texas Presented at: American Association oj Drilling Engineers Drilling Fluids C01!ference April, 1996 - HOI/stOll, Texas The Denick® Hi "G"TM Dryer was the first shaker designed specifically for screening cuttings. It has several features which set it apart from a typical linear motion shaker. A typical linear molion shaker has an adjustable bed angle of 0 to +5 degrees, operates at 1800 rpm, produces 4.5 to 50's, and has 3 screen panels. The Derrick Hi "G" Dryer adjusts from 0 to +10 degrees, operates at 1800 rpm, produces 7.3 G's, and has 4 screen panels. The longer screen frame and increased deck angle combine to slow the
cuttings transport, giving more screen time for the cuttings, which allows greater liquid recovery.
Figure 10-1 P1Y~file
ID·I
View a/Derrick Hi "G" Dryer with 10 bbl Sump
~ERRICK ~~~~.,~
EQUIPMENT COMPANY
Chapter 10
Hi
"e"
Dryer
Originally, the Hi "0" Dryer was intended for lise with oil based muds recovering additional liquid phase from the oil coaled cuttings. It was and is still being used successfully in this role. However, there are other less obviolls ways that Hi "0" Dryers can be utilized to reduce drilling costs. The various way in which savings can
be attained, along with case histories, will be covered in this paper. Although the advantages of using a Hi "G" Dryer are inter-related, they are divided into the following broad categories:
High "G" Dryers Allow Aggressive Screening on Main Shakers Increased Recovery of Liquid Phase
Environmental Protection
High "G" Dryers Allow Aggressive Screening on Main Shakers It is always advantageous to screen as fine as possible at the shakers. Shakers are the most cost effective piece of solids control equipment available. A Hi "G" Dryer below the main shakers will allow finer screening than normal because it can catch any spillage. To change screens or perform maintenance on one of the shakers, the flow can be shut off and the problem fixed without interrupting or slowing the drilling process because, again, the dryer can catch the spillage during this period of time.
Increasing the efficiency of your solids removal system by only a few percent can mean a substantial savings in mud cost alone. There are other advantages such as reduced well bore problems, higher ROP, and longer bit life, just to name a few, When using some of the more expensive drilling fluids such as pseudo oil based muds, the savings can be tremendolls. As an example, see Case History 1 at the end of this section,
Increased Recovery of Liquid Phase TIle additional length created by the use of the fourth screen panel and the ability to raise the bed angle to + 10 degrees allows more screening time by slowing the cuttings transport. This allows the cuttings to be subjected to the higher G forces longer, thereby reclaiming more liquid phase from the cuttings, The additional G forces are especially helpful when using drilling fluids with high surface tension like some of the pseudo oil based muds.
It is common to use a Hi "G" Dryer with oil based muds to recover the liquid phase from cuttings. Savings are realized froIll this recovered fluid. When using oil based muds, often the overriding reason is environmental compliance. If an expensive but environmentally friendly t1uid is being used, such as pseudo oil based muds, the
C!2
"~O"~-
; DERRICK .-,
__ ,--
£QUIPAlfNTCOMPAI/Y
Chapter 10
Hi HG" DI)'el'
reason for reclaiming as much of the Huid phase as possible is purely economic. An example of an Hi "G" Dryer in an application such as this is presented in Case
History 2.
Environmental Protection Rcducting the percent of oil on cuttings when using oil based muds is of prime importance in many regions. The Hi "G" Dryer has been very successful when used to process cuttings from the rigs main shakers. Environmental restrictions vary from one area to the other. In some areas, the cultings can be buried on site provided that the oil on cuttings value meets certain standards, otherwise the cuttings must be hauled off and disposed of. In other areas, all the clIttings must be hauled off and disposed of properly, regardless of the oil on cuttings valuc. A good example which illustrates reduction of haul off and disposal fees is presented in Case History 3. In most cases it is difficult to put a dollar value to the environmental savings, yet the economic savings alone justify the use of a Hi "G" Dryer. See Case History 4 for an example of substantial economic savings.
Case History 1 Shariah Drilling Program Amoco Sharjah Oil Company has conducted an extensive drilling program in the Sheikdom of Shmjah since early 1994. In a 67 day period of drilling, the company recovered 1,255 bbl of mud, saving $62,750 in mud costs. The net savings for the period totaled $35,815, about $S35/day. No dollar values were placed on environmental benefits. A key to these savings was a Hi "G" Dryer from Derrick Equipment Company, equipped with corrugated pyramid screens. Oil tools International Ltd. supplied the solids control services for Nabors Rig No. 128, including three DC1Tick Model 48 FloLine Plus Cleaners, one DelTick Hi "G" Dryer, and two S2-1 fully hydraulic centlifuges, all serviced by an engineer. The centrifuges can operate either in solids removal or in barite recovery. Because of previous experience and thorough knowledge of geological sequences, Amoco Shmjah typically completes a 12,000-ft well in the area in about 4 weeks. TIle wells are usually drilled with oil-based mud, as was the 16 inch hole section on this wel1.
10-3
Chapter 10
Hi "G" Dryer
Dlyer Routine analysis of equipment parameters, such as throughput rates and separation factors, showed that a substantial amount of mud was discharged with cuttings into the waste pit. To try to eliminate this mud loss, Amoco Shatjah put into service the recently developed Hi "G" Dryer. The dryer has an elongated screen bed and high-G linear-motion separation which are suited to recover residual oil on cuttings at a second separation stage between cuttings discharge and entry into the waste pit. Amoco Shmjah approved a 3-week trial of the Hi "G" Dryer. The unit characteristics include 0.19 inch stroke with two 1800 RPM motors (60 Hz power) producing an 7 ,3 G acceleration force. It can be fitted with four 48 inch x 30 inch screen panels. Special skids were built and explosion-proof augers were installed under the shakers discharge assembly to adapt the dryer to the rig. ll1is work was completed during a 4 day rig move, so the dryer was available at spudding.
Test Well The objectives included the following: Recover cuttings from shale shakers to dry them and optimize free mud recovery; Use finer mesh screens than normal because excessive mud disch;:u·ge could be recovered using the Hi "G" Dryer; Create a "last line of defense" to prevent any mud from spilling over the shaker caused by screen blinding or other unforeseen problems; Reduce any environmental impact from oil discharged into the pit, thus reducing the amount of oil ultimately to be treated or removed; Evaluate the value of recovered mud vs rental and screen costs for the dryer; and Determine if tripping could be reduced through improved hole conditioning and mud make up because of second-stage cLittings cleaning for difficult formation intervals. The results from the 3 week test were encouraging; Amoco then kepl the Denick equipment as a permanent pmt of it.s solids control for the remaining drilling program.
10-4
Hi HG" Dryer
Chapter 10
Since the initial test, additional wells have been drilled with this cleaning system. Table 10-1 lists the economic benefits of this cleaning system.
Table 10-1 Mud Cleaner Economic Data/or Four Wells
Cost, $
Operation Tolal Mud Saved, bbl'"
2,213
Days Drilled
129
Footage Drilled, ft
52,928
Rental Costs Screen Costs
187
(39,210)
85
(18,123)
Gross Savings, Mud
110,650
Net Savings
53,317
Net Savings. $ft drilled
1
*Mud Cost $50/bbl
From June 17, 1994, to Jan. 3,1995
Case History 2 Chevron Nigeria, First Hi "G" Run Chevron recently conducted its first Hi "G" run, using the Hi "0" dryer to reclaim oil from cuttings discharged from the main shakers. Chevron was very pleased with the performance and economic savings which was achieved with the equipment. They were especially pleased that they were able to utilize a finer mesh screen on the rig shakers. This provides mare efficient solids removal and results in a lower solids content in the active mud system. The hole section was 12 1/4", with 11.0 ppg oil content at 85-87 GMsIKg, 8.5% by weight. Chevron used a formula that gives the average loss of Illud per foot drilled.
I O~5
Hi "C" Dryer
Chapter 10
Table 10-2 Average Mud Loss hy Screen Size Mesh Screen PMD [25 PMD [40
PMD [75 PMD 175
Bb[s [osUFt.
.058 .08 .67
.57
The mud engineer reports that the figure may be slightly lower, but use of the Derrick equipment over the last three 24 hour periods resulted in a savings of 10 to 15 Bbis of oil mud recovery per 24 hours. This means that the averllge basic savings to Chevron ( 10 BbIs x $350/Bbl.) = $3,500.00/day. Chevron engineers and staff were all impressed, as the Hi "0" exceeded their performance expectations.
Case History 3 Shell Oil Company, McAllen Ranch/Lips Field Study
A recent study was done by Tom Uhlfich & John Simon of Shell Oil Company in South Texas to evaluate the performance of the Derrick Hi "G" Dryer. The study was done on several wells on the McAllen Ranch and the Lips Field just north of the McAllen Ranch in Hidalgo County. In these fields, the oil based cuttings are put into rail car type containers (Millers, Pan American, Cajun) then hauled off (0 a disposal site (Campbel1s). The wells in Lips Field did not lise the Hi "0" Dryer. The wells on the McAllen Ranch did use the Hi "0" Dryer. An average cost per well for previous McAllen wells is listed [or comparison. Shell Oil found that the Hi "0" Dryer not only paid for itself in oil based mud savings, it also substantially reduced haul off and disposal costs by 35-40%. In the tables below, the "Campbells" cost represents costs for oil based cuttings disposaL Based on the savings achieved, Shell Oil is recommending using the Hi "0" Dryer on al1 their domestic oil based mud jobs.
10-6
Hi "G" Dryer
Chapter 10
Table 10-3 Past McAllen Wells Average Cost Pool Water based land farm on site Millers Campbells
$ 1,800.00
$ 4,000.00 $36,203.00 $ 29.152.00 $71,255.00
Total
*Pool - Vacuum Truck Company *Millers - Environmental Company - Railroad Cars & Haul off *Campbells - Oilbase Cuttings Disposal Site
Table 10-4 Well Costs Using Hi "G "M Dryer Gild Cajlln COllstruction McAllen 90 Total Depth = 12,170' Total Days:::: 32
Pool Water based land farm on site Cajun:l: & Dryer Campbells Total Oil Base Recovery 320 BBLS Total
@
$50.00
$ 2,090.00 $ 4.000.00 $ 37.250.00 $ 18.428.00 $ 61.768.00 ·(16.000.00 $ 45,768.00
lVIcAllen 91 Total Depth = 12,720' Tota! Days = 44 Pool Water based land farm on site Cajun* & Dryer Campbells Total
Oil Base Recovery 320 BBLS Total
@
$50.00
$ 1.800.00 $ 4.000.00 $ 37.250.00 $ 14,892.10 $ 57.942.10 ·(16.650.00 $ 41,292,10
*Cajlln Construction - Turn key haul off (Trucks, Tanks, & Backhoe)
10· 7
Chapter 10
Hi "G" Dryer
Case History 4 SEDCO FOREX Project, Solids Control System Upgrade In November 1994, Diltools (Europe) Limited was awarded a contract to upgrade the Solids Control systems on both the SEDCO 711 and SOVEREIGN EXPLORER for the Brittania Project SEDCO FOREX. It required a system that would meet the strict Oil on Cuttings (DOC) disposal criteria ofIess than 80 GMs/Kg, and be efficient and economical in operation and maintenance. In addition, all solids were required to be slurrified for pumping overboard away from the subsea template. This case history highlights the benefits and quantifies the savings realized on the SOVEREIGN EXPLORER after the first three wells of a nine well program.
The Hi "G" Dryer was chosen due to its proven field performance. It can be utilized as either shaker or cuttings dryer. For the purposes of this contract, two Hi "G" units were utilized: One as a shaker, with a standard sump One as a dryer, with a 1011bl sump The design of the Hi "0" Dryer incorporates a single four panel screen deck with 42.4 square feet of actual screen an~a (utilizing the new Pyramid Plus Screen). The screen deck can be angled up to 10 degrees to further reduce fluid loss. The unit is available with an optional 10bbl slimp tank under the screen which can be utilized for fluid storage plior to transfer.
Desigll Philosophy All four shakers and the four panel Hi"G" Dryer (operating as a shaker) were installed in traditional parallel mode. The second Hi "0" was installed downstream of the shakers to process all cuttings and ensure environmental compliance. The screen underflow from the second Hi "G" Dryer was transferred to the active system or to the DE-IOOO Centrifuge system for fine solids removal processing. The drill cuttings from the second Hi"O" Dryer were then transferred to the Oiltools SllllTification System for eventual disch,uge overboard, away from the subsea templ<.lte.
10-8
Chapter 10
Hi HG" Dryer
Results The Derrick Hi "G" Dryer not only ensured environmental compliance with DTI regulations but exceeded the Operator's own in-house target of 80 GmslKg. A total of 534 bbls of oil based mud was recovered and OOC figures were significantly reduced. It must be noted that the OOC target sct by the Operator was met even though the ll1njolily of screens ulilized on the shakers were 180-230 mesh. This can be directly attributed to Derrick's Pyramid Screen technology which allows greater screening capacities from the same size shaker withollt increasing ooe level. Over the first three wells, fluid losses overboard equated to 0.085 bbls/ft.
Conclusion It must be noted that until the advent of Derrick Hi "Gil technology, the traditional
method for control of ooe was to install a cuttings wash/cleaning system. This was not only an expensive solution, but in some cases a very inefficient one, which sometimes reported higher figures than that being discarded from the shakers! As it is, the Hi "G" Dryer requires minimal .'>upervision. Together with its low operation costs and significant savings in fluid. it is an attractive allernative to any cuttings cleaning system. It can also be utilized well with high cost "pseudo" muds and/or "closed loop" systems. Another major benefit of the Hi "G" Dryer technology, sometimes overlooked, is the elimination of whole mud losses from the shakers. This historically has ensured that the shakers were never run at their optimum capacity and thus the solids control system efficiency was adversely affected. Incorporating the Hi "G" Dryer down stream allows the shakers to be run close to maximum capacity without the concern of whole mud losses.
Equipment Layout
Chapter 11
Figure 11-1 Layout for Dual Centrifuge System IFith Weighted Mud
Return
to ....(11
Hole
n ". "I
Oegasser Flow Line
""us'ab'" eq"ab,e,
no,""",,>,,,,",,
Screened SolidS Discard
11-1
:F=lc§:':::Ln ~
t
!-:t~~,~ Un'''''""'"'
fOu/PMENTCOMPAIiV
Equipment Layout
Chapter 11
Figure 11-2 Layout jor Single-Stage Cel1lrijuge System ,vith Weighted Mud
Return to Hole
...(11
Ad,"".b1. ~c."e<
n"","'Y"'9h
t Screened
Di~~~~
"FE:::J..V
t
Centnfuge
Rud D,"""d
11-2
Equipment Layout
Chapter 11
Figure 11-3 Layout for Single-Stage Centrifuge System with UnIFeighted Mud
Return
to -(u
Hole
11-3
Chapter 12
API RP 13 C
Drilling Fluid Processing System Evaluation Equipment 1. Install water meters on all lines that will supply water to the mud system. The meters shall comply with American Waterworks Association Standard C705. Metering of oil should be done in accordance with API Manual of Petroleum Measurement Standards. 2. Mud weight (density) determination - use a Illud balance. 3. Water, oil, and solids determination - ro, 20, or 50 cc retort (preferably 20, or 50 ee). 4. Chloride lest kit appropriate for water or oil based tluids.
Special Considerations L Do not use water on centrifugal pump bearings or packing. 2. On offshore rigs, make sure the seacock line can be isolated from the active mud system so that no ull-metered water is introduced to the system.
Procedure Whatever sampling method used should be consistent because different methods can result in inaccurate comparisons. 1. Measure and record all suction pit mud weight. salinity, and solids (retort) data for subject interval. 2. Determine the base fluid fraction (Fbt') and drilled solids fraction. We recommend that you do the base fluid fraction when the hole is relatively free of cuttings. such as from casing seat to casing seat. Begin drilling out a casing seat uncilhe hole is clean, then again after this hole section is drilled and circulated clean prior to running the next casing strings. a. The base fluid fraction (Fhf) is the percent of liquid in the drilling fluid (water & oil), corrected for salt. b. The drilled solids fraction is the percent of solids in the drilling fluid, cOlTected for salt. Use the solids analysis calculations provided at the end of this section,
12-1
APIRPI3C
Chapter 12
Calculation 12-1 Ca/cillationsfor Solids Removal System Pelfonnance 1. Volume of mud built ~ Vmh
:::::
Vmh = volume of mud built Vhf = volume of base t1uid added
Fur::::: base fluid fraction 2. Calculate the excavated volume of solids drilled - VIIs. Vd~
= (hole diameter 2 / 1029.4)(footage)(1 + washout fraction) (20% = .2)
This gives volume in bhIs.
3. Calculate the total dilution (Dt). Total dilution is the amount of mud that would be built if there was no solids removal equipment
Dt = Vds=
Total dilution
F{l~=
Drilled solids fraction. To calculate, use Solids Analysis method to calculate the low gravity solids (see the follow-
Excavated volume of solids dril1ed
ing page), This also includes commercially added products such as bentonite, thinners, etc. You must account forthese and subtract from the low gravity solids to get your drilled solids. Maintain records and calculate chemical concentration or estimate. Whichever method is used, be sure to maintain consistency for consistent results. 4. Calculate dilution factor (Dr) - Ralio of volume of mud built to total dilution. Ratio of mud used lo actually drill an interval using a solids removal system as compared to only using dilution.
5. Calculate drilled solids removal system performance. SP = (1 - Dr) (100) 12-2
API RP 13 C
Chapter 12
Table 12-1 Solids Ana(-vsi... Formulas
A = chlorides in PPM B = % water
C =%oil
o
= Mud weight (ppg)
E = Specific gravity of weighting agent (API Barite = 4.2)
= Specific gravity of drilled solids (typically 2.6)
F
G = Specific gravity of oil (diesel is ,84) H =Bx( I
+ (5.88 x 10-8 ) x A ")
= 100-C-H K = I +(1.94x 1O-6)xA 9 5 = (( 12 x D) - (H x K)-(G x C» II
L
= % volume of salt water = % volume corrccted solids = specific gravity of salt water = average specific gravity of solids
M = (I x (E-L)) 1 (E - 2.6)
= % volume low gravity solids
N =I-M
= % volume high gravity solids
o
= N x E x 3.5028
= high gravity solids in ppb
P
= M x (2.6 + F ) x 3.5028/2
= low gravity solids in ppb
,r. Average specifIc gravity is typically between 2.2 & 4.2. It could be higher jfthe
specific gravity of the weighting material is higher, such as with hematite or galena
J 2-3
Chapter 12
API RP I3 C
Example of Solids Analysis A B C D E F
20000 ppm 80%
11% 9.6 ppg ~
G H
[
K
L
M
N
0 p ~
4.2 2.6 .84 B x (l + (5.88 X 10. 8 ) x A 12) 80 x (l + (.000000059) x 20000 12) 80 x (I + (.000000059 x 144955.9327) 80 (I + 0.0085524) 80 (1.0085524) 80.684192 = % volume of salt water 100-11-80.68 = 8.32 = % volume cOITected solids I + ( 1.94 X 10'6) x 20000. 95 I +0.00000194 x [2189.29242 I + 0.023647 1.0236 = Specific gravity of salt water ((12 x 9.6) - (80.68 x 1.0236) - (.84 x II» 1 8.32 (l15.2 - 82.584 - 9.24) 1 8.32 23.376/8.32 2.8096 = Average specific gravity of solids (8.32 (4.2 - 2.8096)) 1 (4.2 - 2.6) 8.32 x 1.3904) 11.6 7.23 = % volume low gravity solids 8.32 - 7.23 1.09 = % volume high gravity solids 1.09 x 4.2 x 3.5028 16.0358 = High gravity solids in ppb. 7.23 x (1.6 + 2.6) x 3.5028/2 131.691268/2 65.845:::;; Low gravity solids in ppb.
12-4
API RP 13 C
Chapter 12
Example of Solids Removal System Performance Vhf Fbf
~
4000 bbls .97
V",
900 bbls
Eh
Total low gravity solids - commercially added low gravity solids 7.23 (from solids analysis) - 3 (3% commercially added products) 4.23% Drilled solids or .0423
Volume of mud built VJl\h = ~
VblJFbf
40001.97
~4123.71
Total Dilution
= Vds I Fels ~ ~
Df
900 1 .0423 21276 bbls.
= Vmb / Dr ~4123.71
~.193819
System Performance
SP ~
12-5
( j - Dr) 100 (I-.193819)( 1(0) (.80618) 100 80.62 %
121276
bbls built
Chapter 12
API RP 13 C
Notes on the API RP-13 C Calculation API 13C makes no allowances for the use of old mud. It is intended for calculation in a section of the well from casing seat to casing seat. Start with new mud in the section of the well where the calculations are to be made. If stored mud is used, then you have to treat all the low-gravity solids (additives and drilled solids) as commercially added products. Make sure YOli keep good enough records to do proper mass balance at the end of the section (API does not address mass balancing). 1. This mud analysis is done after circulating the hole clean prior to pulling out of the hole and running casing. Remember that the % volume aflow gravity solids is the amoLlnt of drilling solids plus commercially added products (gel, caustic, thinners, polymers, etc.). 2. For accuracy, all additions of waler, oil, and chemicals should be accounted for. The volume of fluids added is the volume of water and oil, if any. The base fluid fraction is the percentage of fluid. To calculate the percentages, one must account for the amount of commerically added materiaL 3. If you know the excavation solids (volume of hole drilled), you can input it, or input the hole geometry by pressing calculate, input the data and press next. 4. Input the amount of commercially added product in percent. Either use a mass balance or keep all inventory of the volume of chemicals and additives used. The total dilution figure is the amount of mud that would have to be used to dilute back to the desired percentage of drilled solids ifthere was no solids removal equipment used. 5. This is the drilled solids removal system performance for the input values.
12-6
API RP I3 C
Chapter 12
&
"'
:g
0
~
"'
0
~
"' "' 0
~
0
~
0 0 '<:t
0
14) (") .,~-
c-f
0 0
r 0
(")
0
14)
N
0 0 N
..a ..a
o-~;
0
,f
f'
~
.;:: 0
o~~ 0
'"E
t;--
:::J
"0
;:se
> .- 0'" 0
0
r
14)
I
'. ,=" c,..-
0 0
0
..... •...
0
14)
0 14)
0 .....
12· 7
..... 0
Il"l
CI)
CI)
Mud Weight, Ib/gal
14)
cO
[
How To Optimize Solids Control Economics, Efficiency by Leon Robinson, Ph.D. Solids control efficiency can be Il10re effectively managed using calculations that include desired removal efficiency, the amount of drilled solids in the Illud discard stream, the targeted drilled solids concentration and the balTels of drilling fluid that must be added for each bane! of solids discarded. Many years ago, a controversy raged about the effect of drilled solids on well costs. Many thought drilled solids were an inexpensive substitute for weighting agents. As drilling encountered more difficult environments, however, hole problems finally became undeniably associated with excessive drilled solids. Most rigs now process all drilling fluid sequentially in accordance with good practices. l.~ The drilling fluid type does not affect proper rig plumbing. Dispersed or nondispersed, fresh or salt water, clay-base or polymer-base or any other drilling tluid must be treated sequentially to remove smaller and smaller drilled solids. Solids control equipment has historically been economically justified as an insurance policy for catastrophe prevention. Today's more expensive drilling fluids also require low drilled solids concentrations. Polymers that adhere to active solids are expensive and require significantly lower concentrations of drilled solids than a Iignosulfonate drilling fluid. Environmental concerns also dictate tight control over the mechanical removal of drilled solids to minimize waste fluid. The following discussion shows: How dilution is necessary. How too much dilution creates problems with excessive drilling fluid volumes and excessive drilling fluid wastes. How removal efficiencies affect the drilling fluid costs. How the dryness of the drilled solids affects drilling fluid and drilling waste costs. Three primary variables - dryness of discarded solids, targeted drilled solids concentration in the drilling fluid and removal efficiency are important in evaluating a rig solids management system. They all playa role in minimizing drilling fluid and drilling waste costs. The dryness of discarded solids varies significantly with different lypes of solids control equipment. Orbital motion shale shakers with coarse mesh screens discard very dry solids (60%-80% by volume solids), but also pass a large fraction of solids presented. Linear motion shakers with fine mesh screens can remove more solids, but more liquid drilling fluid also is discarded. The fluid removed with these fine mesh screens may have 20%-40% by volume solids and contain more or less weighting agent than the fluid in the pilS.
Hydrocyclones settle solids inside of the cones in the whole drilling fluid. Spray discharge is desirable, but seldom do all cones on a drilling rig discharge solids in a spray discharge. The solids concentration in hydrocyclone discharges can vary from 20%-40% by volume. The centrifuge underflow usually has a solids concentration between 55%-65% by volume and the overflow has a solids concenh'ation between 5%-15% by volume. The mud cleaner screen discharge usually resembles the concentration of the centrifuge undertlow. Thus, a value around 35% by volume solids concentration of the overall system discard is probably typical, but obviously not specific. DJilled solids concentration targeted in the fluid is a function of the type of dJilling fluid, Polymeric additives that adhere to active solids will be attracted to most clays. These drilling fluids demand low drilled solids concentrations. Inhibitive fluids also need low-solids fluids, but they also aid the removal processes by not allowing the clay-type of solids to disperse. The advantages of low solids drilling f1uid have been well documented. The removal efficiency of the solids control system cun VUly signific,mtly depending upon adherence to sound UlTangement practices, an operating conditions, The new proposed API RP BC plans to provide a new method of following a tracer to detennine the solids removal efficiency. A second method is also presented here. These methods can be used while dlilling a reasonably long interval ( 1,000-2,000 ft). The analysis assumes thal no drilling fluid is jetted from the system. Removal efficiency in an unweighted drilling fluid is usually higher than in a weighted drilling fluid. Solids in the same size range as the weighting agent cannot be removed at this time. Rigs with poor solids management have trouble reducing the mud weight below 9.7Ib/gal. other rigs can comfortably create an 8.8lb/gal drilling fluid. This is related to removal efficiency. Dilution One way to keep drilled solids at a manageable level is to simply replace drilling fluids containing solids with clean fluid. Half of the drilled solids can be eliminated if half of the system is dumped and replaced with clean fluid. This usually is too expensive, so mechanical equipment is used. If the mechanical equipment does not remove a significant portion of the drilled solids reporting to the surface, dilution becomes critical to the solids management strategy. Fig. 1 shows how the quantity of new drilling fluid depends upon the effectiveness of the solids removal equipment. The calculations assume that the Fig. 1. The total volume of new drilling fluid is 286 bbl, determined by the volume needed to fill the pits (281 bbl) less the volume of solids returned (0 bbl) plus excess volume to dilute (0 bbl). The discard volume is also 286 bbl, determined by the volume removed (286 bbl) plus excess volume built (0 bbl) to dilute solids, In example 1b, the total volume of new drilling fluid is 254 bbl, determined by the volume needed to fill the pits (257 bbl), less the volume of solids returned (10 bbl) plus excess volume to dilute (7 bbl). The discard volume is 286 bbl, determined by the volume removed (257 bbl) plus excess volume build (0 bbl) to dilute solids. In example 1c, the total volume of new drilling fluid is 480 bbl, determined by the volume needed to fill the pits (229 bbl) less the volume of solids returned (20 bbl) plus excess volume to dilute (271 bbl). This discard volume is 500 bbl, determined by the volume removed (229 bbl) plus excess volume build (500 bbls less 229 bbl) to dilute solids.
2
FATE AND EFFECT OF 100 BBl OF DRY SOLIDS REPORTING TO THE SURFACE 35% vol solids discarded 4% targeted drilled solids concentration Fig. 1 a. 100% solids removal efficiency FATE
-B } B
Step 1 Volume discarded = 1.QQJ;lbl = 286 bbl 0.35 Step 2 No dilution required
EFFECT ON PIT VOLUMES
286 bbl
~;i~~::
P - Step 1 pit level
Slep 1
Volume discarded = 286 bbl.
286 bbl required 10 retum pils 10 original level
286 bbl
New drilling fluid required = 286 bbl
Step 2
Fig. 1 b. 90% solids removal efficiency FATE
,
,
EFFECT ON PIT VOLUMES
257 bbl
B
-
~riginal_B 250
pit level - Slep 1 pit level
Step 1
Step 1 Volume discarded = 90 bbl = 257 bbl 0.35 Step 2 Dilution volume = 10 bbl = 250 bbl 0.04 Total volume discarded = 257 bbl.
New drilling fluid required = dilution volume ~ solids returned + additional fluid to fill pits = 250 bbl-l0 bbl + 7 bbl Step 2 = 247 bbl Fig. 1c. 80% scUds removal efficiency bbl}
7 bbl addilional fluid Dilution volume
FATE
Step 1 Volume discarded = .8..O....QQL = 229 bbl I
B
Ii
Step 2 0.35 Dilution volume = 20 bbl = 500 bbl 0.04 Total volume discarded = 229 bbl. + 271 bbl = 500 bbl
EFFECT ON PIT VOLUMES O'i,iOO'_
229 bbl
Step 1
pi.llevel - Step 1 pillevel
271 bbl 229 bbl
Excess fluid buill }
New drifling fluid required = dilution volume" solids returned = 500 bbl - 20 bbl = 480 bbl
229 bbl required 10 return pits to original level
Step 2
3
solids control equipment is removing either 100%, 90% or 80% by volume of drilled solids. The average drilled solids concentration in the discard stream has been selected as 35% by volume. The undertlow from a decanting centrifuge and the mud cleaner discharge stream will contain 35%-63% by volume of solids. This example assumes that the sUlface system contains 1,000 bbl of drilling fluid and the targeted drilled solids level is 4% by volume. It also assumes that 100 bbl of drilled solids report to the surface. For reference, 100 bbl is the volume of 1,000 ft of a 10-in. diameter hole. This analysis is the basis of a minimum desirable removal efficiency to build the minimum quantity of new drilling fluid. This also minimizes the volume of discarded fluid. In a 100% removal efficiency case, all drilled solids reaching the surface are removed along with 63% by volume ofthe dlilling fluid. Pit levels will decrease by the volume of dIilled solids and drilling fluid removed from the system, assuming that the solids compressibility is zero (Fig.1a). No drilling fluid is needed to dilute the drilled solids retumed to the system. However, 286 bbl of drilling fluid is needed to keep the pit levels constant. The drilled solids concentration will decrease. In other words, a 4% volume drilled solids level will not and cannot be sustained. To achieve a 90% removal efficiency, the volume of drilled solids and drilling fluid removed is 257 bbl (Fig.lb). The solids returned to the system (10 bbl) must be
EFFECT OF REMOVAL EFFICIENCY ON THE VOLUME OF NEW DRILLING FLUID REQUIRED
6 u
•• • -~
~
n·" ~
~£
n 2
8%
5 10%
0
0
~
,
4 I
--:f2'o7~
I 3
,I " "
~
<1>£
Et ,0
-
0
o.
2
> "
..
~~ 0:::: 0
"
0
o 40
50
60
70
80
90
100
% solids removal effioienoy barrels of solids removed per barrel of
solids reponing to the surfaoe.
Fig. 2 The chart can be used to define the volume of drilling fluid that must be added for each barrel of drilled solids reporting to the surlace, assuming that discharged drilled solids have a concentration of 35% by volume.
4
diluted to 4% volume concentration with 240 bbl of new drilling fluid. This means that an additional 7 bbl of new dri11ing fluid must be built to keep the pit levels constant. This is almost a balanced system If the discarded volume exactly matches the required volume needed for dilution. The minimum quantity of drilling fluid will be built. The optimum removal efficiency for any targeted drilled solids level may be calculated by mathematically equating the removal volume to the dilution volume required. For 80% removal efficiency, 229 bbl of drilled solids and drilling tluid will be discharged (Fig Ie). Although this is only 21 bblless than the 90% removal efficiency, the dilution volumes me significantly higher. Diluting 20 bbl of returned drilled solids to a 4% volume level requires adding 480 bbl. The reconstituted 500 bbl of drilling fluid will contain 20 bbl of drilled solids and 480 bbl of clean drilling fluid, Since only 229 bbl of space is available, 271 bbl of drilling fluid must be discarded to keep the pit levels constant. The total discard will therefore be the 229 bbl from the solids removal equipment and the 271 bbl of drilling fluid, The minimum discm'd volume will occur when the system is balancecl- no excess fluid is needed to dilute the drilled solids returning to the system. The same solids removal efficiency that provides the minimum quantity of new dri11ing fluid to be built will also be the removal efficiency that generates the minimum discard volume. DRILLING FLUID REQUIRED IF DISCARD STREAM AVERAGES 25% BY VOLUME
O~~-L~
50
55
60
__
65
~~-L~
70
75
80
__~~_
85
90
95
100
% solids removal efficiency barrels of solids removed per barrel of
solids reporting 10 the surface,
Fig. 3. The minimum quantity of drilling fluid that must be added if the drilled solids concentration is 25% rather than 35% is slightly larger.
j
Optimum Removal Efficiencies The three examples illustrate that an optimum removal efficiency, around 90%, exists for the 4% by volume target concentration of dIilled solids. A series of calculations can define the volume of new drilling fluid required per ban'el of dIilled solids reporting to the surface (Fig. 2). These calculations assume the aggregate discharged drilled solids have a concentration of 35% by volume. If the removal efficiency is 75% and the targeted drilled solids concentration is 4% by volume, over 6 bbl of dlilling fluid is required for each batTel of drilled solids reporting to the surface. This concept can be extrapolated to other concenu'ations of drilled solids using the same calculational procedure. For example, with 6%-volume targeted drilled solids in the fluid and a 60% removal efficiency, Fig. 2 shows that over 6.2 bbl of are needed for every barrel of drilled solids reaching the surface, As the efficiency improves, the fluid requirement decreases, At 70%, slightly more than 4.5 bbl is required per barrel of solids; at 80%, slightly more than 3 bbl are needed; and at 85%, almost 2.5 bbl are required, For removal efficiencies greater than 85%, the pit level decreases more than the additional new fluid required to dilute the solids to the % by volume leveL The additional fluid required to keep the pit volumes constant will dilute the solids to a value lower than 6% voL
REMOVAL EFFICIENCY WITH 40% VOLUME OF DRILLED SOLIDS IN THE DISCARD
...
2
~
:~2%
10% 18%
6%
o 50
55
60
65
70
75
80
85
90
95
100
% solids removal efficiency barrels of solids removed per barrel of solids reporting to the surface.
Fig, 4, In the 40% volume scenario, a removal efficiency of 80% and a target concentration of 4% drilled solids would require about 5 bbl or drilling fluid for every bbl of solids reporting to the surface,
6
If the drilled solids concentration in the discard is 25% instead of 35%, the minimum quantity of drilling fluid required is a little larger (Fig. 3). The volume of new drilling fluid required for each ban'el of discarded drilled solids are shown in Fig, 4 and Fig. 3 for 40% and 30% volume of discarded drilled solids. These curv~s have a shape similar to those already discussed.
Economic Analysis Drilling fluid costs vary greatly. For a particular well situation, the cost ot;.fl~lid associated with each barrel of discarded drilled solids can ,be ~alculate~ from the information just presented. For example, using the data in fig, 5,_ if the dri.JHng fluid target solids is 4% and the system has a 60% removal effici~ncy, 9.,6 bl;:t.1 of drining fluid will be discarded with each balTel of drilled solids discarded. For a $~O/bbl tiuid, 1,000 bbl of drilled solids would require $480,000 for new drilling fluid. If the drilling fluid target solids can be increased to 6% by volume anc.! .the removal efficiency increased to 70%, only 4.7 ban'els of drilling fluid must be,bu~.lfor each barrel of discm'ded drilled solids. For a $501bbl fluid, the same 1000 bbl of drilled solids discarded would cost $235,000. These charts can be used to evaluate the cost effectiveness of various solids control equipment. This analysis will be related only to the cost of drilling fluid. It will not account for the benefits of reduced trouble-costs or reduced discard volumes.
DRILLING FLUID REQUIRED IF DISCARD STREAM AVERAGES 30% BY VOLUME
55
60
65
70
75
80
85
90
95
100
% solids removal efficiency barrels of solids removed per barrel of solids reporting to the surface.
Fig. 5. In the 30% by volume drilled solids in the discard scenario, a removal efficiency of 75% and a 4% target concentration of drilled solids would require 6 bbl of drilling fluid for every bbl of solids reporting to the surface,
7
Waste Volumes When waste volumes are calculated for various drilled solids concentrations and removal efficiencies; the minimum volume of waste occurs at the same removal efficiency value as required to drill a well (Fig. 6). A mechanistic model describes_all operations ..1 Steady state, dtill-ahead operations with active mechanisms for drilling tluid dilution, cuttings removal, dewatering and recycling are described alone with drilling process discontinuities. The mathematical description is a comprehensive process description that can handle all operations from spud to total depth. EcologicaIJy-driven drilling fluid and cuttings management focuses on reducing drilling tluid mld waste disposal costs. The benefits of minimizing drilling waste disposal volumes by minimizing hole washouts throughjudicioLis selection of drilling fluid, drilling, smaller diameter holes and correct use of solids control equipment has been documented:! ll1ese discussions highlight the dilemma of using salts for well bore stability. Though salts decrease drilled-solids disposal volume, they create disposal problems and additional costs. Determining Removal Efficiency If a well is spudded with fresh water and pit levels remain constant during drilling, solids returned to the system will increase the density. This analysis assumes that no whole drilling tluid is dumped from the system or lost downhole. As the well gets deeper mud weight will increase at different rates for different removal efficiencies.
DISPOSAL VOLUMES
4%\
\1,
~
~
•
~ •0 _t 0 ~ ~
.•
4
B
3
i\
,,[\
~~
~~
c
~
w.s Et
.2~
o. > "
2
ro·
.~
1---1---1----1 -
0:=
0
:
0
<5••
0 40
_ _L_ _
50
60
70
80
90
100
"'0 solids removal etficiency barrels of solids removed per barrel of solids -reporting to the surface.
Fig. 6. Waste volumes are calculated for various drilled solids concentrations and removal efficiencies.
8
METHOD TO DETERMINE
REMllVA~
EFFICIENC1(,.
10.5 10.3
d'
10.1
,1.'1',
il' P. "
.
300
350
400
Hole volume drilled bbl
,
Fig. 7. As a well gets deeper, mud weight will increase at different rates fOT different removal efficiencies.
ESTIMATING SOLIDS REMOVAL EFFICIENCY
•E, g
Volume of_ hole drillBd bbl
Fig. 8. This technique can be used to help analyze solids control economics for any drilling fluid density.
9
Of'
As an example, assume that the surface volume is 1,000 bbl and 2% by volume bentonite is added to the ne~fdr~lli~~ flui.d~ If~t~e Il19d1Weight is 9,3 Ib/gal after drilling 200 bbl afhale, the remov
"' !,
.,
..
', /
,.'.
I'
,
·'I,i,_J/.'o'i!
'-,;-,
10
.',
1-', •
.Refer~nces
'.,1"-;
, ,- ,ri
:i,,' ;._l:i:i
. _. ~ i •
t. Young. Grant and Robinson, Leon: "How to Design a Mud System foi" Optii-iilim Solids Removal," World Oil (September 1982) 57-61; (October 1983) 105-110; (November 1982) 159-174. 2. On11Sby. George: "Correction of Common ElTOfs in Drilled Solids Removal Systems." paper-presented at the Second Adriatic Symposium on Oil Well Drilling, Porec, Yugoslavia (May 7-10,1973). 3. Wojtanowicz, AK.: "Dry Drilling Location - An Ultimate Source Reduction Challenge: Theory, Design, and Economics," paper SPE 260 13 presented at SPEI EPA Exploration & Production Environmental Conference, San Antonio, Texas (March 7-10, 1993). 4, Carter, Thomas S.: "Reduce Drilling Waste Disposal Costs," Petroleum Engineer International (September 1993) 56-61. 5. Walker, T.O. and Simpson, l.R.: "Dtilling Mud Selection for Onshore Operations" Ocean Ind. (October 1989) 43-47. 6. Edward-Berry, J., Robinson, L.H., and Annis, M.R.: "Pycnometer Accurately Determines Low-Gravity Solids," Petroleum Eng;neer International (June 1991 ).
ABOUT THE AUTHOR Leon Robinson is principal and CER of International Drilling Consultants based in Houston. He recently retired from Exxon after 39 years of specialization in drilling research. He has a doctorate in engineering physics from North Carolina State University. Robinson also is involved with the MIT Steering Committee for Rapid Excavation, Mining, Tunneling and Drilling, lhe Houston Chapter of the AADE; the AADE/AFS Committee on Shale Shakers, API 13C workgroup, the standing conmlittee of the API and AADE: and the Houston Museum of Natural Science Energy Exhibit.
II
