Chapter 52
Mud Logging Alun H. Whittaker,
Exploration Logging IX.*
Introduction Conventional mud logging has been comm ercially available since 1939 . The service involves involves ex traction o gases from the returning mud stream an d analysis analysis of the gas for combustible combustible hydrocarbons. Commonly, the resulting an alyses are logged at drilled dep th and plotted alongside a drill-time drill-time or rate of penetration log and a cuttings sample geological log. Although the mud log data cannot be related directly to undisturbed reservoir p roperties, they are important indicators of potentially productive horizo ns in the well. The conventional mud log continues to be the most important geological data source available before wireline log s are run. The mud logging unit offers a useful location for the operation of other wellsite analyses and services. It provides a clean, well-lighted laboratory area with a stable electrical sup ply and is continuously oper ated by geologists or geologically trained technicians. Many mud logging contractors have made use of these assets to augment con ventional mud logging w ith an extensive range of geological and engineering services. Often unrelated to the traditional gas analysis function of the unit, these services nevertheless generally are considered aspects of mud logging now in the same manner as sonic, density, and neutron neutron logs often are group ed w ith “electric logs.” The earliest expansion of mud logging serv ices began in the 196 0’s with the introduc introduction tion of improved improved methods of geopressure detection. detection. New techniques techniques were added to the logging logging unit and and it became common for a separate “pressure log” log” to be prepared alongside the mud log. The 24-hour activity of the mud logging unit allowed continuous operation of this service service in which early detection was essential. In the 1970’s, the advent of rugged microelectronics allowed the introducti introduction on of more sophisticated sophisticated and autom ated equipmen t into the logging unit. Most notably, the use of drilling drilling rig data-acquisition systems ‘The chapter
on this this topic In the 1962 edibon was written by A.J. Pearson.
linked to minicom puters introduced a range of drilling drilling optimization and control services. U nlike conventional mud logging logging and geopressure detection, detection, these services services are essentially essentially nongeological. Generally, engineering personnel personnel are added to the logging logging crew for these In the 1980’s three new aspects to mud logging services have been introduced. First, direct links links between wellsite minicom puter and an office data center allow centralized surveillance and control of several wells. The logging unit provides a wellsite access point to the central com puter data base and and analytical softw are. Second , th ere is the increasing increasing use of the mud logging unit as the surface receiving and control center fo downhole measurement-while measurement-while-dril -drilling ling (MW D) services. The mud logging unit provides b oth a convenient work measurem ent) for this service. Additionally, the abilit to integrate integrate mud logging and MW D data in a single single com puter adds economy and speed to the well evaluat evaluation ion process. Third, the 1980’s have brought the first fundamental changes in the methods of hydrocation hydrocation and geological geological analysis, which continue to be the comm on denom inator of all mud logging servic es. Impro ved sampling techniques, pyrolysis, pyrolysis, chromatography, and other geochemical techniques have enhanced the diagnostic and quantitative value of mud logging. logging. Wellsire geoche mical screening for reservoir and source-bed type may now be performed in the mud logging unit.
Service Types The number and range of mud logging logging contractors is possibly greater than that of any other oilfield oilfield service. The logging services offered by any single contractor may range from basic hydrocarbon logging, using equipment barely barely mo re sophisticated sophisticated than that introduced introduced 40 years ago, to complex chemical and physical physical analyses analyses and a com plete engineering surveillance and control center.
52-2
PETROLEUM
ENGINEERING
HANDBOOK
Similarly, logging personnel may be gradu ate geologists or engineers, or technicians of various levels of expertise. In specifying the mud log service for a well, the opera tor’s engineer, geologist, and the logging contractor should define the objectives and problem s anticipated ticipated and select those aspects aspects of the service service required. Since extra service usually implies extra cost,
of maximum mud flow rate. Ports in the lower part of the trap allow mud to enter and leave the trap. An electric or pneumatic agitator motor provides provides both pumping pumping and degassing of mud passing through the trap. Gas evolved evolved from the mud is mixed w ith ambient ambient air in the upper part of the trap and drawn through a vacuum line to the logging unit for analysis. This device pro vides
ing monitoring, it is relatively easy to comp ute the saving in drilling drilling time or cost required to justify some addition to logging service day rate. This is discussed in detail at the end of this chapter. Although overlaps occur, the services provided in mud logging may be grouped in line with the traditional oilfield oilfield disciplines: (1) formation evaluation services ’ -hydrocarbo n analysis, geological analysis, and geoche mical analysis; (2) petroleum engineering services-geop ressure evaluation and petrophy sical measurem ents; (3) drilling drilling en@ neering services-data acquisition and data analysis. This order is convenient for the following discussion of logging services since i closely parallels the historical developm ent of mud logging and the level of sophistication of logging units used today.
continuous gas sample. How ever, the efficiency of the device is som ewha t affec ted by drilling drilling p ractice. Pum rate and ditch mud level will influence influence mud flow rate through the trap; trap; mud rheology will be a factor in the the degassing efficiency of the trap; and mud and ambient air tempe rature around the trap and vacuum line will affect the relative relative efficiency with which light and heavy hydrocarbons are extracted and retained retained in the gas phase. This latter effect is most noticeable in areas of high diurnal temper ature variation, whe re heavier alkane gases seen in daylight may condense and be lost in in the cold of night. An alternative alternative to the conventional conventional gas trap is is the steam, or vacuum, mud still. In this device, a small sample of drilling drilling mud is collected at the ditch, returned to the logging unit, and distilled u nder vacuum. The meth od provides a relatively high and uniform extraction efficiency for all hydrocarb ons. It is, how ever, a timeconsuming manual process. A nalyses are noncontinuous and subject to human error; fo r exam ple, light hydrocarbons can evaporate while the sample sample stands prior t analysis. While a useful addition to the conventional gas trap at times, the mud still does not provide a real alternative. The development development of a continuous continuous gas trap with good and consistent efficiency of extraction is a high priority in the improvement of mud logging logging technology.
Formation Evaluation Services Gas Extraction Extraction Methods
Although the modem mud logging unit may perform many different services, probably its most critical one is the analysis analysis of hydrocarb on gases. 2 Befo re this analysis analysis can be performed, a sample of gas must be extracted extracted from the drilling drilling mud. This is perfor med by the gas trap (Fig. 52.1). The ga s trap is a square o r cylindrical metal b ox immerse d in the shale shaker ditch, preferably in a location
Fig. Cl-Gas
extraction
at
the ditch
52-2
PETROLEUM
ENGINEERING
HANDBOOK
Similarly, logging personnel may be gradu ate geologists or engineers, or technicians of various levels of expertise. In specifying the mud log service for a well, the opera tor’s engineer, geologist, and the logging contractor should define the objectives and problem s anticipated ticipated and select those aspects aspects of the service service required. Since extra service usually implies extra cost,
of maximum mud flow rate. Ports in the lower part of the trap allow mud to enter and leave the trap. An electric or pneumatic agitator motor provides provides both pumping pumping and degassing of mud passing through the trap. Gas evolved evolved from the mud is mixed w ith ambient ambient air in the upper part of the trap and drawn through a vacuum line to the logging unit for analysis. This device pro vides
ing monitoring, it is relatively easy to comp ute the saving in drilling drilling time or cost required to justify some addition to logging service day rate. This is discussed in detail at the end of this chapter. Although overlaps occur, the services provided in mud logging may be grouped in line with the traditional oilfield oilfield disciplines: (1) formation evaluation services ’ -hydrocarbo n analysis, geological analysis, and geoche mical analysis; (2) petroleum engineering services-geop ressure evaluation and petrophy sical measurem ents; (3) drilling drilling en@ neering services-data acquisition and data analysis. This order is convenient for the following discussion of logging services since i closely parallels the historical developm ent of mud logging and the level of sophistication of logging units used today.
continuous gas sample. How ever, the efficiency of the device is som ewha t affec ted by drilling drilling p ractice. Pum rate and ditch mud level will influence influence mud flow rate through the trap; trap; mud rheology will be a factor in the the degassing efficiency of the trap; and mud and ambient air tempe rature around the trap and vacuum line will affect the relative relative efficiency with which light and heavy hydrocarbons are extracted and retained retained in the gas phase. This latter effect is most noticeable in areas of high diurnal temper ature variation, whe re heavier alkane gases seen in daylight may condense and be lost in in the cold of night. An alternative alternative to the conventional conventional gas trap is is the steam, or vacuum, mud still. In this device, a small sample of drilling drilling mud is collected at the ditch, returned to the logging unit, and distilled u nder vacuum. The meth od provides a relatively high and uniform extraction efficiency for all hydrocarb ons. It is, how ever, a timeconsuming manual process. A nalyses are noncontinuous and subject to human error; fo r exam ple, light hydrocarbons can evaporate while the sample sample stands prior t analysis. While a useful addition to the conventional gas trap at times, the mud still does not provide a real alternative. The development development of a continuous continuous gas trap with good and consistent efficiency of extraction is a high priority in the improvement of mud logging logging technology.
Formation Evaluation Services Gas Extraction Extraction Methods
Although the modem mud logging unit may perform many different services, probably its most critical one is the analysis analysis of hydrocarb on gases. 2 Befo re this analysis analysis can be performed, a sample of gas must be extracted extracted from the drilling drilling mud. This is perfor med by the gas trap (Fig. 52.1). The ga s trap is a square o r cylindrical metal b ox immerse d in the shale shaker ditch, preferably in a location
Fig. Cl-Gas
extraction
at
the ditch
MUD LOGGING
Hydrocarbon
52-3
Analysis
PRESSURE :VACuUM UMP 1 REGULATOR
The basic form of gas analysis analysis involves the analysis by combustion combustion of the bulk sample. A lthough commonly called “total gas analysis,” analysis,” it is, in reality, analysis for total combustible gases and primarily detects the lowmolecular-w eight alkanes (paraffins) such as methane ethane, pro pane, butane, and pentane (with partial concentrations of hexane and heptane at higher ambient temperatures). Catalytic Combustion Detector (CCD). After filtration and drying, the gas stream stream is injected at constant pressure and flowrate into into a detector chamber (Fig. 52.2). The original original type of mud logging gas detector, and probably still the most widely used, is the catalytic combustion, combustion, or “hot wire” detector (Fig. (Fig. 52.3). The hot wire detector is a Wh eatstone bridge circuit consisting of four resistances: resistances: a fixed resistor, Rf; rheosta t, R ,, used to trim or balance the bridge; and a match ed pair of coiled platinum wire filaments, Rd and R,. The two filaments are enclosed in an analysis cell with the detector filament, filament, Rd, exposed to the flow of gas sample sample and the reference filament, R,, isolated in pure air. When a bridge v oltage, V, is applied, the filaments filaments become heated. A voltage between two and three volts is commonly selected to give a high enough filament temperature for hydrocarbon combustion combustion at the filament filament surface (actual voltage used depends on the particular detector design). design). Combustion heat causes the temperature and hence resistance resistance of the detector filament to rise relative to the reference filament. The bridge is unbalanced and current flows between th e two sides of the bridge. Using a galvanom eter of resistance R, , this current, I,, can be measu red. Since combustion occurs at the filament surface only, the galvanom eter current is quite sensitive sensitive and linear with changing gas concentration. Obviously, detecto r response w ill depend on both the concentrati concentration on and composition composition of the sample gas phase, since each hydrocarb on species will have its own particular heat of combustion. Table 52.1 sho ws these for the low-m olecular-weight alkanes. Since gas compo sition is unknown, the total gas detector cannot be calibrated for tme compositional response. The dete ctor is calibrated with a mixture of a single alkane, usually m ethane, in air. Detector response is then repor ted in in percentag e “equivalent methane in air or EM A. Using a variable resistance, R,, in the bridge it is possible to adjust the bridge current, I,, and graduate the galvanometer directly directly in percentage EMA An older practice, practice, which is now becoming obsolete, was to take the galvanom eter reading in milliamps and relabel it as “gas units.” units.” Such units units are obviously equipment specific although some company or regional standards have been enforced. Where this practice continues, continues, confusion can be avoided by requiring requiring the logging contractor to repor t calibration calibration data on the mud log heading. For example, the contractor contractor would report “ 100 total gas units=2% EMA.” Fig. 52.4 shows the response of a typical typical CCD to commonly occurring combustible gases. Notice that response of 1% EMA, or 50 total total gas units, units, may indicat a concentrati concentration on of 1% methane or a somewhat lower
Fig. 52.2-Gas
analysis system.
ZERO
ADJUST
POTENTIOMETER
L
SPAN
ADJUST
POTENTIOMETER
Fig. 52.3-Catalytic
TABLE 52.1-HEATS
Cn
+2n+2)
combusti on detector.
OF COMBUSTION ALKANES
(3n + 1) +-O,-nCO,+(n+l)
n=1
Molecular Weight
OF THE SIMPLE
H,O+E
(kcallmol)
kcallgm
Methane
1
16
191
11.9
Ethane
2
30
342
11.4
Propane
3
44
493
11.2
Iso-butane
4
56
648
11.2
Butane
4
58
650
11.2
Structure
u” tit
j-c+4
52-4
PETROLEUM
ENGINEERING
HANDBOOK
+ i20
HYDROGEN
SAMPLE
+1
7+
+-J
Fig. 52.5-Flame
ZERO AO.JvsT
0
0.2
0.4
0.6
0.8
PERCENT
Fig. 52.4-Catalytic
1.0
x2
““ D R O t .a R B O N
I.4
16
, 8
20
IN AIR
combusti on detector response.
concentration of a mixture of methane and heavier alkanes. Total gas response may be thought of as a “gas richness indicator,” increasing both with gas concentration and with addition of heavier fractions. To assist in discriminating light alkanes from heavy ones, a second identical detecto r m ay be used. By setting a lower bridge voltage (1 to 1.4 V) and filament tempe rature, the detecto r is no longer capable of inducing combustion of methan e. The resulting de tector output, still repor ted in percentag e EM A, is comm only labeled petroleum vapors, wet gas, or heavies, although only qualitative comparison of the two detector responses allows recognition of dry and oil-associated gas shows. Response of the CCD can be maintained linearly up t the stoichiome tric, or ideal, combustion composition of hydrocarb on in air. Above this compo sition, appro ximately 9.5% EM A, the detector “saturates,” incomplete combustion occurs, and response becomes nonlinear. At higher concentrations, the sample must be diluted with air before it is introduced into the sample cham ber to maintain a combustible gas mixture. Theoretically, by using progre ssive sample dilution, a
*I ionization detector.
thermal conductivity of the gas mixture an d induce a small heating ef fect at the detecto r filament. This positive response is comm only so small as to be insignificant when com pared with the greater hydrocarbon response. How ever, if the concentration of noncombustible gases becomes so high as to prevent complete combustion of hydrocarb on with air, a much larg er negative response will occur. Hydro gen will bum in the detecto r, even at low voltage, giving a concentration response similar to methane. Although free hydrogen does occur as an in termed iate produ ct of petroleum maturation, it is extremely reactive and diffusive. Occurr ence of hydrog en in a petroleum gas show is therefore most uncommon. Significant co ncentrations of hydrog en ha ve been shown to result from deep-seated structural movement, but the most common origin is from the corrosion of aluminum drillpipe or of steel drillpipe in extreme ly low pH drilling fluids. A serious disadvantage of the CCD is the tendency of the catalyst surface to become poisoned by the accumulation of impurities and partial combustion products. This may result in a slow, pro gressive degradation of performance or a sudden, catastrophic loss, when, for exam ple, silicon compou nds are present in the mud. Regular detecto r calibration is essential to maintain reliable operation.
100 % EM A. In reality, ea ch dilution stage requires a reduction in gas sample volume and an increasing mixing error. It is generally accepted that 40% EMA is the maxim um limit of reliability of a CCD As an alternative to progressiv e dilution, whe re high gas concentrations are regularly expe cted, the detecto can be reconfigured to operate as a thermal conductivit detector (TCD). Though the detector circuit remains essentially unchanged, it is opera ted at a lower voltage such that no gas combustion occurs. B ridge current now is reversed, responding to the cooling effect of the gas stream passing over the detector filament. Methane, which ha s a substantially greater therm al conductivity than air, will produ ce a large cooling effect, w hich m ay be linearly calibrated up to very high concentrations. The device is, however, poorly responsive to the heavier alkanes, CO2 and hydrogen sulfide (HzS), which have therma l conductivities close to that of air. The high conductivity g ases, hydrogen and helium, will give responses even greater than that of methane. The CCD is quite selective for hydrocarbons. Carbon
Flame Ionization Detec tor. The inherent limitations of the CCD resulted in a search for a more reliable detector technology . The most accep ted and increasingly used is the flame ionization, or “hydrogen flame,” detecto (FID) (Fig. 52.5). One important differen ce between th e flame ionization and the catalytic co mbustion principles is that the flame ionization m ethod involves complete comb ustion of the sample. A small quantity o f sample is introduced into hydroge n/oxygen mixture that is continuously burning in a combustion chamber. The heat generated by the hydrog en flame is sufficient to initiate com plete co mbustion of all hydroca rbons in the sample. A large oxygen exces s is maintained relative to the small sample volume and saturation never occurs. The heat output o f the hydrogen flame is the sum of the heats of combustion of hydrogen and the sample hydrocarbons. Unfortunately, most of the heat produced is from the large volume flow of pure hydrogen. The small, dilute flow of hydrocarbons produces such a small proportion of the total heat of combustion that it cannot be measured accurately. Combu stion heat then cannot b e used as
at the detecto r filament. They will marginally reduce the
hydrocarb ons
instead
relies
upon
an
unusual
in-
MUD LOGGING
termed iate stage in combustion that only occurs in hydrocarb ons burning at high temp eratures. This involves the creation o f unstable electrically charg ed anions and cations. By placing a positive electro de, or anode, in the form of a cylindrical chimney above the hydroge n flame, th e negative anions may be collected and the resultant e lectric current used to determine hydrocarbon concentration. The ionization/combustion sequence is a comp lex one that involves many interm ediate and alternate reaction steps. The number of ions created, and therefore the current flowing, is in direct prop ortion to the concentration of the alkanes and to the number of carbon atoms in the alkanes (F ig. 52.6). The FID response in percentag EMA is, therefore, like the CCD , a richness indicator showing increases with increasing concentration and increasing alkane molecular weight. The FID is totally selective for compo unds containing carbon-to-hydrogen (C-H) bonds. Other gases and impurities in the sample stream p roduc e zero or negligible response and do not degrade detector performance. Althoug h th e detecto r respon se is effectively linear through out all concentrations, the electrom eter used to monitor and amplify the detector current has performance limits of linearity. Since mud log gas show s may vary fro m tens of parts per million (ppm ) to tens of percent, both electrical signal attenuation and sample splitting are required to ensure low -range sensitivity and high-range linearity of FID response. In most mod em instruments this is handled autom atically, ensuring a higher degree o f accuracy than manual sample dilution. Gas Chromatography. In addition to a total gas detector, most mod em logging units will also contain a gas chromatograph. This device allows the separation of the individual alkanes and their separa ted detection, giving a gas analysis of compo sition and concentration. While this analysis is of greater value than th e total gas response in EMA , the chromatograph does not provide a continuous analysis but proce sses batch samp les separate d by a number of minutes. In drilling terms, this translates into separa te analyses several feet apart. The chromatograph does not replace the total gas recorder in showing the fine detail and progressiv e changes in a gas show. In gas chromatography, a fixed volume gas sample is carried through a separating column by a carrier gas, usually air. The column contains liquid solvent surfac or a fine molecular sieve solid. By difference in gas solubility or by differential diffusion, the gas mixture becomes separated into its components, the lightes traveling most quickly through th e column and the heaviest most slowly. Depending on the nature of the column, each component will pass throug h and exit the column in characteristic time. From the column, the components pass in turn to a detector, which may be a CCD , TCD, or FID. The detector is calibrated with a gas mixture of known compo sition and concentrations. A separa te calibration factor for each component can be used for detector response as the components occur in turn. Since heavier com ponents take longer to traverse the column, the time and depth interval betw een sam ples is governed by the number of components to be analyzed.
52-5
0
0.2
0:4
HYDROCARB ON
Fig. 52.6-Flame
0.6 CONCENTRATION
0.8
1 :o
% IN AIR
ionization detector response.
In routine logging, a chrom atogra ph usually will be set to cycle throug h continuous automa tic analyses f or methan e, ethane, propane, isobutane and n-butane. This requires ap proxim ately 3 to 5 minutes. If heavier alkan es (e.g., pentanes) need to be detecte d, th e automa tic control is disengage d and the analysis allowed to continue for a longer period of time. Infrared Absorption Detector. The third, and least used, form of detector is the infrared absorption detector. This instrument uses the principle that any chemical bond will absorb infrared energy o f a specific frequ ency governed by the chemical nature and geometry of that bond. Fo r exam ple, methane contains four identical carbon-to-hydro gen (C-H ) bonds. If a gas sample is irradiated with infrared energy at a frequency characteristic of this bond, the energy absorbed by the sample will be in proportion to the number of C-H bonds and hence to the concentration of methane in the sample. All other alkanes contain C-H and carbon-to-carbon (C-C) bonds. Althou gh these bonds are chemically identical, they vary in geom etry and hence characteristic infrared frequency, depending on their position within the alkane molecule. Theoretically it should be possible to pass the gas sample through a series of test cells, testing for infrared ab sorption at a series of characteristic infrared frequencies. Combination of the results would provide a continuous analysis of both alkane type and concentration-i.e., the equivalent of a continuous chromatogmph. Unfortunately, the C-H and C-C bonds show such a large number o f minutely varying ge ometr ies th at, instead of a series of discrete c haracteristic frequencies, continuous band of overlapping absorptions occurs. At best, using a two-absorp tion cell system, it is possible to provide an estimate of methan e concentration and total hydrocarb on concentration, in EM A. This result is comparable to the result obtainable from a dual CCD system and inferior to the results from an FID-equ ipped gas chromatograph.
The most Detection of Nonhydrocarbon Gases. comm only occurring nonhydrocarbo n gases in petroleum exploration are CO*, HzS, helium, nitrogen, and hydrogen. As discussed previously, the occurrence of naturally prod uced h ydrogen is rare. Helium and nitrogen also tend to have regionally or geologically
PETROLEUM
52-6
specific occurrences. CO2 and H2 S arc common trace or significant compon ents of natural gases and equipment for their detection should be used on any exploration well. Chrom atograph-Ther mal Conductivity Detector The most versatile device fo r detection of nonhydrocarbons is a chromatograph equipped with a TCD. By selecting an appro priate column material and length, any single compon ent or combination may be separa ted. The TCD will provide a response to any gas that has a thermal conductivity different from that of the carrier gas. This response will differ for each sample gas/carrier gas combination but, by using gas mixtures of known composition, calibration curves for each compon ent can be developed. Best response and sensitivity is achieved when the maximum difference exists between the thermal conductivities of the sample gas and the carrier gas. Thermal conductivity generally declines exponentially with increasing molecular weigh t. Thu s the light gases (hydrogen and helium) may be readily detected by use of air as the carrier gas. For the heavier gases (nitrogen, CO;, , and HzS) th at have thermal conductivities closer to that of air, a lighter and higher thermal conductivity carrier gas must be used. Helium is a common choice but hydrogen also may be used if available. An important co nsideration when assessing the reliability of analyses for nitrogen and CO2 is their presence in the sample caused by the introduction of air into the gas trap and the aeration of drilling fluid. Air of normal atmo spheric compo sition, dissolved and entrained in drilling fluid, will be introduced continuously into the boreho le. In the hot downh ole environment, corrosion and other oxidation reactions will deplete oxygen from this air resulting in a relative increase in Oxidation of carnitrogen and CO* concentration. bonaceous material will further add to CO2 enrichment. Alternatively, the presence o f corrosion inhibitors in the mud may deplete both oxygen and CO*. Regardless of the mechanism involved, oxygen depletion will increase with tempe rature and length of circulation time through the downhole system. At the surface, this oxygen -depleted air and any gas recovered from the formation is mixed with ambient air at the gas trap. This air will vary in composition with the emissions surrounding atmosphere-e.g., from rig motors, vehicles, and others. Any show of nitrogen or CO2 from the formation must be recognized above backgrou nd concentrations, which will show som e random variation and a progressiv e increase as the hole is deepen ed and mud circulation becomes hotter and of longer duration. Regardless of the analytical meth od used, p recision of the ppm level cannot be provided by the analysis. When only trace quantities of gas are expected or when a precise compositional analysis is required, mud logging analysis of CO or nitrogen cannot b e relied upon.
Of the nonhydrocarbon petroleum gases, HIS and CO? are the most significant. They are the most commonly occurring gases in high concentrations and because of their polar nature pose serious problems of corrosion of drilling and production equipment. H2 S is
ENGINEERING
HANDBOOK
also toxic in relatively low concentrations. In many areas, detectors specific to these gases are considered standard mud logging equipment. Infrared Absorption Detector. Continuous CO2 detection is best handled with infrared absorption. An infrared analyzer is used that is responsive to the characteristic frequency o f the carbon-to-oxygen (C-O ) bond unique to COz. Correction of atmospheric CO2 concentration is perfor med by alternately scanning tw sample cells. One contains a sample from the gas trap and the other contains ambient air. Differential output provides measure of CO2 concentration above atmospheric. Tube-Type Detector. Several types of H2S detectors are used, all of which m onitor a change resulting from the chemical oxidation of the gas. The simplest detector is the tube-type device in which th e sample gas mixture is drawn at a controlled flow rate through a glass tube containing reactive lead acetate. The lead acetate, w hich is deposited on a substrate of high-surfac e-area silica gel granules, reacts with HzS to produce lead sulfide and changes from white to dark brown or black in the process:
Pb(CH3C00)2
+H2S-‘2CH3COOH+PbS.
Since the amount of lead acetate in any unit length of tube is constant, the tube may be graduated in terms of concentration of H2S in a fixed volume of sample. The panel-mounted instrument has two tubes installed. Flow of sample from the ditch is constant through one of the tubes, and, if H2 S is present in the gas being evolved at the ditch, the lead acetate begins to discolor progressively from bottom to top (the direction of sample flow). Since the sample, and hence the discoloration, is this response is qualitative only. T he continuous, discoloration indicates that H2S is present in only trace or in enriched quantities, but no estimate of actual concentration can be made As soon as this discoloration is seen, a warning must be given since even trace quantities of gas can be dangero us. A quantitative analysis can be mad e by switching flow to the second tube and introducing timed sam ple. In this case, a fixed amount of discoloration occurs and the scale allows reading of the H2S concentration. An alternative configuration for the tube indicator is in a small handbellows, often called a “puffer” or “sniffer,” which can be used to sample the atmosphere in various locations around th e rig. the tube is used it must be replaced . The instrument cannot keep a continuous record of HzS concen tration but only a series of individual measurem ents. This is draw back, but not a serious one since any quantity of H 2 S in the atmosphere is both a health hazard and an indication th at the mud system is totally satur ated. Once HzS is detected, mud treatment to remove it must begin Gas measurement is required to ensure that it is removed and does not reappear. The tube indicator may be used to detect CO 2 or any other gas for which a discoloring reactant is available. For CO*, hydrazine is used in place of lead acetate.
52-7
MUD LOGGING
Presence of CO2 is indicated by a purple coloration of the chemical in the tube: CO2 +N2H4 -‘NH2NHCOOH. The tube method, however, is poorly suited to continuous monitoring since there will be a uniform rate of discoloration by atmospheric CO2 Paper-Tape-Type Detector. A more sophisticated version of this detection principle uses continuous paper tape, im pregnated with lead acetate, to allow continuous analysis and a quantitative electrical o utput fo r chart recording and activation of alarms. The detector mechanism is similar in appeara nce to an open-reel tape record er. Its operation and operating components are analogous to that of tape recording. Paper tape, a porou s filter paper co ated w ith an even concentration of lead acetate, is wound from reel to reel at a constant speed. The tape passes through a sample cham ber through which gas from the ditch passes continuously. The tape will be discolored by an amount proportional to the concentration of lead acetate on the tape, the speed at which the tape is moving (both of which are constant), and the concentration of H2S in the sample. From the sample cell the tape passes to a detector where light from a collimated source is reflected from the tape to a photoe lectric cell. The output of the photoe lectric cell is readily calibrated in terms of H2 S concentrations by passing through the system test paper strips with zones of different color that correspond to a range of known concentrations. The paper-tape-type detector m ay be used for the detection of COz or other gases if a suitably impregnated paper tape is available. Unlike th e tube indicator, it is possible to discriminate between a baseline of atmospheric discoloration and a true “show” above baseline. Although the paper-tape-type is superior to the tube indicator, both suffer the disadvantage of requiring periodic rep lacement of the reactive material, lead acetate, and the possible degradation of the product in storage . Ind icator tubes and rolls of paper tap e are supplied in sealed, dated packages and should never be used if the seal is broken or the packa ge is beyond its expiration date. Solid-State Electrical Detector. The most modem Hz analyzers involve use of a solid-state electrical detecto r. This device depends on the reversible reduction of metallic oxides by HzS as its means of detection. semiconductor sensor element is exposed to a flow of gas drawn from the ditch. The surface of this element consists of a proprietary metallic oxide layer. In the presence of HzS , this layer will be partly r educed to metallic sulfides, and its electrical resistance will change:
(metal) O+HzS-+(m etal)
S+HzO
This is an equilibrium reaction. If Hz S ceases to be present, the reaction reverses w ith the reoxidation of sulfides to oxides. At all times, the sulfide-to-oxide ratio (and
hence the electrical resistance) of the layer is a direct function of the concentration of H2 S in the sample p resent in the vicinity of the sensor. Alternative configurations of this device involve multiple installations with either samples being drawn from , or sensor elements located in, various locations around the rig with centralized monitoring and alarm functions. Locating the sensor in a remo te location m ay cause problem s if the sensor is expo sed to potential damage or mistreatment. It does, however, remove the risk of loss of response resulting from ga s dissolving in condensation in long vacuum lines. The device has high reliability and accuracy and i widely used in the industry. There are, howeve r, two deficiencies that should always be considered. The first and most important is that if the sensor is operated for a period of time w ithout any H2S present, it tends to lose reaction speed . (It is important to note here that the sensor does not lose sensitivity! It will respond, within calibration, to the presence of H2S , but will respond somewhat sluggishly to the first appearance.) For safety reasons, the sensor m ust be reactivated regularly by using a sample of H2S to maintain its reaction speed. Second , th e sensor will respond to certain org anic sulfides that may be present in oil or result from mud additive decomposition. The response to these compounds is low but m ay result in a false H 2 S show. Soluble Sulfide Analyzer. One disadvantage common to all Hz S gas analyzers results from the high solubility of the gas in water. H2.S will not be liberated from the drilling fluid and w ill not be seen by a gas analyzer until a saturated solution of the gas exists. Since serious c orrosion problem s may be caused by low concentrations of the gas in mud and even a few ppm of the gas in air is a health hazard, it can be seen that by the time that HzS gas is detected at the surface, a major problem already has developed. Early d etection of H2 S requites analysis of the drilling mud. This can be accom plished by regular sampling and wet chemical analysis, but the mud logging service can provide continuous soluble sulfide analysis by using a selective ion electrode measurem ent system. With this device, a sensor probe, which is immersed in the drilling fluid, contains pH (hydrogen ion) and pS (sulfide ion) specific electrodes and a tempe rature sensor. When HzS dissolves in water it will in part dissociate into bisulfide (HS -) ions and sulfide (S ~ -) ions. The solubility of H2 S and the degr ee of dissociation are controlled by the pH and temperature of the solution If these two parameters and the concentration of single dissolved sulfide spe cies are measured it is possible to deduc e the concentration of all other spec ies. In the soluble sulfide analyzer this is done automatically by a micro processo r. By using this device it is possible to detect H 2 S and begin treatment to remove it from the mud without concentrations ever becoming high enough for gas detectors to be effective or for personnel to be placed at risk. Geological Analysis
After gas analysis, the most important function of mud logging is the sampling and evaluation of drill cuttings.
52-a
PETROLEUM
Even wh en the mud logging unit operator is not a professional geologist, the minimum requirement is for identification and brief description of sample lithology, estimation of reservoir properties (amount and type of porosity and permeability), and description of oil staining. Hydrocarbon and geological Sample Lag Time. analysis depen d on the representative sampling of drill cuttings and gases liberated by the cutting action of the drillbit. In interpreting the analytical results, it is necessary to account for the lag time and physical effects of the gas and cu ttings travel from the bottom of the hole to the surface. ’ Lagging of samples is essential so that results may be reported or logged at the depth from which the sample originated (at the time the sample arrives at surface, the depth, o f course, will be somew hat greater). Lag time may be obtained simply by calculating the time necessary to displace the total annular volume of drilling fluid as given by van=v,
-VP,
v a n =-
,..,.....,.....,...........(l)
.
.
.
.
. .
.
.
.
.
.
.
.
.
.
. .
.
.
.
.
. .
.
.
.
.
.
.
(2)
an
I/=-,
VCWI
...
.. . .......... ..... .
where V = annular volume, m3/m, VT = hole capacity, m3/m, VP = pipe capacity and displacement, = annular velocity, m/s, = pump output, m3/s, qP tl = lag time, s, and D = depth, m.
. .(3)
m3/m,
Separate calculations must be performed for each an nular section (drillpipe in casing, drillpipe in open hole, drill collars in open hole, etc.). Calculated lag times ate used when first drilling out of casing or in hard ro ck areas whe re an in-gauge hole is expected. However, a calculated lag time cannot take int account cap acity variation in out-of-gau ge holes or variation in pump rate or efficiency (for exam ple, when the pump is stopped to make a connection). Determining and using lag in terms of pump strokes has distinct ad vantages over lag determined on a time basis. The counters tracking the cuttings up the hole stop automatically when the pump is stopped. Clocks would continue to run, and some subtractive factor would have to be introduced. The most important advantage, how ever, lies in accuracy . A lag determined in terms of an interval of time is correct fo r only one speed of the circulating pump (that speed at which the lag determina tion was run), whe reas the lag in pump cycles is accurate for any pump rate.
ENGINEERING
HANDBOOK
The lag can be determined by placing a tracer in the drillpipe at the surface when the kelly bushing is “broken off,” allowing the tracer to be pumped through the hole and back to the surface, and counting the number of strokes requited of the circulating pump to mak e this circulation. From th is total pump stroke count, the number of strokes required to pump the tracer down through the pipe to the bottom of the hole is subtracted. This figure is calculated on the basis of the capacity of the drillpipe and the displacement of the circulating pump . The result is the “lag stroke.” Various materials (such as who le oats, barley, or strips of colored cellophane) may be used as tracers and picked up on the shaker screen for approximating the lag. Under ordinary circumstances, how ever, calcium carbide placed in the drillpipe will react with the mud to form acetylene. This gas will be picked up by the mud gas detecto r and is the most convenient and reliable m ethod for determining the lag. Acetylene gas appe ars as et gas on the gas detector and is easily distinguished from methane produced from the formation. There is no Representative Cuttings Samples. substitute for representative cuttings samples accurately correlated to the depth from which they came. They are the required supportive data for the evaluation of any mud logging, geo logical, geophy sical, or engineering data. Every rig has a shaker screen for separating the cut tings from the mud as they reach the surface. The shaker screen may or may not e a good place from which to take cuttings samples. 3-5 If the shaker screen is used, board or catching box should be placed at the foot of the screen for collecting composite samples. This becomes especially important wh ere drill rate is low, to ensure that the sample collected is representative of the whole interval drilled and not just the final few inches. shaker is used, difWhere a traditional “rhumba” ferences in flow through the possum belly (ditch at the rear of the shale shake r) w ill result in density and size sortings of cuttings a cross the various screens. This sorting can be of assistance to the logging geologist in partially sep arating large cavings from the smaller bottomhole cuttings. However, great care must be taken to ensure that a representative sample is caught. Where a modem “doubledeck” shaker is used, cuttings on both the upper and lower screens should be sampled. A sampling depth interval should be set that thz mud logger can be expec ted to maintain while keeping u with other responsibilities. Samp le intervals can be shortened as the hole is deepen ed and drill rate falls. The mud logger should never allow m ore than 15 minutes to pass between catching samples. For example, if the sample interval is 10 ff and the drill rate is 1 0 ft/hr, the mu logger should take four scoops of samples over the hour to fill the sample bag for the interval. Special samples should always be taken whenever background gas change s are seen or the lag time after drilling breaks occur. If a board or catcher b ox is used, it must be cleaned off after each sample is taken. Samples should be taken from the desilter or desander outlets whe never these are running. In this way, the logging geologist can establish th e quantity and appeara nce system.
If an unconsolidated
formation
is penetrated
MUD LOGGING
sample fro m the desander will contain both formation sand and mud solids. The logging geologist must be able to discriminate between these Washing and preparing the cuttings to be examined are probably as important as the examination itself. In hard rock areas, the cuttings are usually quite easily cleaned, in which case washing is a matter of merely hosing the sample in a container of water to remove the mud film. Washing the cuttings in many areas, howeve r, partitularly areas and zones of tertiary sands and shales, is more difficult and requires several precautions. The clays and shale s prese nt are often so ft and of a consjstency which goes into solution and makes mud. Care must be taken to wash away as little of the shale as possible, and, in determining the sample co mposition, to take int account that which is washed away. After washing the cuttings to remove the mud, they are washed through a 5-mm sieve unless doing so will further cause exce ssive loss of shale or clay. It is generally considered that the cuttings will pass through the S-mm sieve, and that the material that does not is cavings and may be discarded. However, the material that does not pass through should be examined for sand cuttings. If they should be present, these afford an excellent opportunity for study of larger-than-normal cuttings chips. Cuttings from wells drilled with oil-based or oilemulsion muds are usually m ore representative of the drilled formation than cuttings drilled with water-based mud because the oil emulsion prevents sloughing and dispersion of clays and shales into the mud. A t the same time, washin g and handling cuttings drilled with this type mud poses somewhat of a problem; they cannot be cleaned by washing in wate r alone. It is usually necessary to wash th e cuttings first in a detergent solution to remove the mud. Some of the liquid commercial detergen ts available may be used. In extrem e cases, it may be necessary to wash the cuttings first with a nonfluorescent solvent such as naptha, and then wash them in a detergent solution to remove the solvent. Use of a solvent is not advisable unless absolutely necessary because of the risk of removing any oil staining present. An oven mounted on the wall of the logging unit can be used to dry a portion of the cuttings sample after it has been washed, but some of the washed cuttings are examined wet under the microscope. A sample of unwash ed cuttings also is required for cuttings analysis in the blender. Althoug h these cuttings sho uld not be rigorously wash ed, a light rinsing to remov e su rface drilling mud film is advisable. The logging geologist should extract a small amount of sample from each stage of the sample preparation process. From examination of all samples, an cc urate estimate of sample composition can be produced. 3 Once the percenta ges of the various constituents have been estimated , the sample description in logical order should contain (1) rock type, (2) color, (3) hardne ss (induration), (4 ) grain size, (5 ) grain shap e, (6) sorting, (7) luster, (8) cementation or matrix, (9) structure, (10) porosity, (1 1) accessor ies, and (12) inclusions. Only a visual samp le examination usually is required at the wellsite in elastic (sand/shale) formations. In carbonates, other tests may be required to determine the chemica l an d physical nature of the rock. Th e simplest of these is to test cuttings with dilute hydroch loric acid; the
52-9
rate of reaction, which is rapid for calcite and slow for dolom ite, provide s a guide to relative com position. This test can be made m ore quantitative by use of calcimeter. In this device a weigh ed sample is treated with acid in a sealed reaction cham ber. Reaction is monitored by measuring either the volume or pressure of evolved CO;! over time until reaction is comp lete. Output is percentag e of calcite and dolom ite in the rock sample. In more complex mixed carbonates and sulfates (e.g., anhydrite and gypsum ), a chemical stain kit may be used. S mall sample s of wash ed drill cuttings are spottested with a series of chemica l test solutions. Chara cteristic coloration of a test solution is indicative of the presence of a particular mineral in the samp le. Many excellent texts are available that discuss th geologica aspects of d logging. These include Low , Maher, 7 and McNeal. * Since this chapter deals primarily with the technology of mud logging, they are not discussed further here.
Hydrocarbon
Content of Samples
In addition to a geological evaluation, cuttings samples must be tested for hydrocarb on content. A blender or cuttings g as analysis must be perfor med on every sam ple caugh t. This involves disintegration of a sample of cuttings in a blender, extraction of a sample of liberated gas, and injection into a total gas analyzer. This can be perfor med by manual extraction with syringe and injection into the unit’s online gas analyzer. How ever, for speed and continuity of operation, mod em logging units use automa tic extraction and injection into an independent catalytic combu stion cuttings ga s analyzer. As soon as a representative cuttings sample has been taken, a measured amount (100 cm3 in a measuring cup) of unwash ed sample is placed in the blender jar and covered by 600 cm3 of water, then blended for 30 seconds and left to stand for another 30 seconds before taking a gas sample. If the hole is caving badly, the amount of cuttings may be increased but should be consistent-especially before and through a show. With hard carbonates, low-po rosity sandstones, or similar reservoir rocks that cannot be efficiently pulverized by the cutter blades (40 to 60 seconds’ blending is recommended), the blender jar should be allowed to stand for 2 or 3 minutes before taking any gas readings. After the gas analysis is performed, the water should be inspected for oil signs or petroleum odor. Any droplets can be skimmed off for examination. The crushed rock material also may be of value in clarifying lithological evaluation. The blender is a good evaluation tool because it gives some indication of the quality o f the reservoir w ith respect to the porosity and the GOR. A good porosity sandstone generally will be well-flushed by the time it reaches the surface, so the amount of cuttings gas obtained will increase pro portionately to the decreasing porosity. This is also true with a sucrosic do lomite or high-porosity limestone such as chalk. How ever, if the reservoir is a fractured carbonate, etc., with all the oi and/or ga in the fractures, little or no cuttings gas will be recorded and the use of the blender as a porosity in dicator is of limited value, bec ause fu ture production is
PETROLEUM
52-10
going to be more dependent on the complexity of the reservoir fracturing than the inherent po rosity and permeability of the rock itself. In oil reservoirs, gas is normally in solution with the oil, and the agitation of a cover ed sample provides an excellent index of the amount of gas with the oil, wh ich is significant in view of the gas already rec orde d from the ditch. A high cuttings gas with an oil show should be treated as a very significant show and sho uld be one of the more important factors to consider in the overall evaluation. When large intervals of reservoir rock are cored, the blender readings obtained are not likely to be as informative as those obtained if the section ha d been drilled normally. Generally, the amount o f sample is reduce because the center is still in the core barrel. With the often slower drill rate, the percentag e of cavings may be increased. Also, if a diamond head is being used, the rock will be coming back in a very ground-up and often badly altered state. Thus, if the geologist is agreeable representative loo-cm3 samples from the more brokenup parts of the recovered cores may be blended with wate r, and any readings can be used to supplemen t the readings obtained during the actual coring. Inspection for liquid hydrocarbon s should be mad e at the micro scope (oil-stained cuttings), the blender jar (petroleum sheen and odor after blending), and in an ultraviolet light inspection box (fluorescent oil droplets on cuttings and diluted mud samples). Visible oil stain and color is an important indication of oil presence and type as are ultraviolet fluorescence intensity and color, grading from dull brown for the heaviest (residual or wet) oil to bright blue-white for light oils and condensate. How ever, crosscheck ing of observations is essential to confirm the presence of oil. Many mud additives, contaminants, minerals, or rig floor debris will have an oily appearanc e or odor and may fluoresce under ultraviolet light. Only if visible stain and ultraviolet fluorescence yield the same conclusion is oil confirmed . For exam ple, samples contaminated with pipe-dope will have a dark oily ap-
Fig. 52.7-Comparative
results of REII, OSA, and THA
ENGINEERING
HANDBOOK
pearance suggesting heavy oil. Howev er, the same cuttings examined under ultraviolet light will show a bright blue-white fluorescence characteristic of the highest gravity. This incompatability allows the identification of a contaminant and avoids th e logging of a false show. A good mud logger should examine all mud additives stocke d at the wellsite and determine, before their use, their characteristic properties and appearance w hen mixed w ith drilling mud or cuttings. If a true oil stain is identified, a single, representative cutting should be tested with an organic solvent. This is the “cut test.” Solvent cut is valuable in assessing fluorescence and allows deductions to be made of oil mobility and permeability of the reservoir. By removing the oil from the colored background of the cutting, the solvent allows a better estimate o f fluorescence. The way in which the solvent cut occurs (e.g., instantly for highgravity oils, more slowly for more viscous lower-gravity oils, or irregularly streaming from limited permeability) also yields u seful inform ation. If no cut can be obtained from a washed cutting, the test should be repeated on a dried cutting, a crushed cutting, or after application of dilute hydrochloric acid. This will produ ce the required cut and yield further evidence on permeability or effective porosity. After the cut solvent ha s evapo rated, residue o f oil remains in the cut dish, displaying the oil’s natural color. Finally, if sufficient oil is present, it may be possible to determine its refractive index. Just as oil stain color and fluorescence progressively change w ith oil type, refractive index correlates well with oil gravity. Portable refractometers that require only a small droplet of sample are available for use in the mud logging unit. By using a small quantity of oil (skimmed from the surface of the blender jar or a diluted mud sample) , a very reliable estimate of oil gravity can be obtained. Geochemical
Analysis
More sophisticated analyses of hydrocarbon and hydrocarb on source material involve th e principle of pyrolysis-thermal decom position of a sample in an inert atmo sphere . Three such devices are presently available: the Rock-Eva1 II*” (RE), the Oil Show Analyzer’” (OSA), and the Therrnalytic Hydrocarbon Analyzer’” (THA ). All these devices use variations of the lnstitut FranGais du Pitrole-Centre de Recherches du Groupe Petrofina (IFP-FINA) temperature-program ed pyroanalysis method developed by Espitalie. 3,9 The process involves the heating o f a weighed rock sample throug h an increasing tempe rature prog ram in an inert helium stream. Since combustion cannot occur, the helium carries away from the sample hydrocarbons and CO 2 resulting fro m the thermal volatilization of petroleum and organic source m aterial in the rock. These evolved gases may be analyzed by flame ionization and therma l conductivity detecto rs. The amount o f gas expressed as milligrams per gram of rock, the evolved gas, and the time and temperature of evolution may be used to characterize the richness and type o f a reservoir or source rock. The differences between the three devices are shown in Fig. 52.7. The RE 11 uses a uniform temperature ramp of 25” Umin up to 550°C. The helium stream carrying evolved gases passes to a CO2 trap and then to an FID.
MUD LOGGING
On completion of the pyrolysis the trapped C O2 is passed to a TCD. The output showing temperature, FID and TCD respo nse is called a “pyrogram.” The RE II pyrogram characteristically shows tw distinct pe aks in FID re sponse. The first, SI , represents true hydrocarb ons, oil and gas, volatilized from the sample. The second, 52, represents hydrocarbons generated by the thermal cracking of hydrogen-r ich organic sou rce material, kerogen , in the sample. The temp erature, maxi at which the peak of S2 occurs is indicative of the maturity of the kerogen. Mature kerogen, capable of generating oil or gas, will have a T,,, in the range of 435 to 470°C. A lower T,,, indicates imm ature kerogen and a T,,, above 470°C indicates postmature material that has already yielded the majority of its hydrocarb on product. The TCD response, S3, represents the yield of CO2 from the thermal cracking of oxygen-rich kerogen in the sample. A comparison of S2 and S3 provides the relative hydrogen /oxygen richness of the kerogen . Th is is useful in estimating source type. Hydrogen-rich kerogen is prone to rich oil yields, whe reas ox ygen richness gives more gas-prone and lower-yield kerogen. The oil show analyzer (OSA) differs from the RE II in that it uses a nonuniform tempe rature consisting of two temperature steps followed by a temperature ramp. Following completion of pyrolysis the sample is further heated in an oxygen atmosphere causing the complete combustion of all remaining organic carbo n in the sample. The OSA pyrogram generally shows three characteristic peaks in FID response with SO and Sl corresponding to the two temperature steps and representing the splitting of the RE II Sl peak into a lowertempe rature (gas-indicating) peak and a highertempe rature (oil-indicating) peak. The S2 peak and T,,, are the same as those seen in RE II. S4 represents CO2 produ ced by pyrolysis (S3 equivalent) and by combustion. Combination of the pyrolysis and combustion gas products provides a measure of the total organic carbon content or the gross organic richness of the rock. RE II has become widely used as a laboratory instru ment and both it and the OSA have seen use in the mud logging unit in frontier exploration. The restriction on their wider implementation in mud logging ha s been the high complexity (and price) of these instruments, which has limited the ir use to the most advan ced logging units and demanding exploration environm ents. The THA, a much simp ler dev ice, is better suited to routine m ud logging se rvices. It uses only an FID and a temperature program similar to the OSA pyrolysis phase (without th e final combustion phase) . The THA pyrogram provides S O, Sl, S2 and T,,, . Neither C O;! analysis, S3, nor S4 is available from the THA.
The Modern Mud Logging Unit There ar six basic requirements for a modem mud logging unit* based on the previous discussions. 1. A total combustible gas analyzer using catalytic combustion or flame ionization detector 2. A gas sample dilution system, allowing maintenance of linear detector response at high gas concentrations or a backup therm al conductivity detecto r. 3. An automatic cycling chro matograph capable of
52-11
isolating
and detecting m ethane, etha ne, propan e, and isobutane or a second, low-voltage catalytic combustion detector , allowing discrimination of “total gas” from “petroleum vapors.” 4. A separate cuttings gas analyzer, allowing g as analysis from blended cu ttings samples. 5. A microsc ope and ultraviolet light inspection cham ber fo r the identification and description of lithology and liquid hydrocarbons. 6. A pum pstroke counter, which , in conjunction with calcium carbide lag tests, allows gas readings and cuttings samples to be lagged back to correct drilled depth. In addition, the unit requires a drilling depth and time recorder fo r the determination of sample depth and the calculation of rate of penetration, an important rock strength/poro sity indicator. Ideally, this should be independent of the driller’s depth recorder. Since mud logging samples (gas, oil, and cuttings) are extracted from the mudstream, changes in mud chemistry and rheology must be considered when evaluating mud log results. The logging unit should be equipped to perform basic mud tests-e.g., mud balance, Mar sh funnel, sand test kit, and filter press. Laboratory glassware and chemicals are required to perform chemical tests and titrations on cuttings and mud filtrate samples. Although pressure control is not a standard function o mud logging (see Petroleum Engineering Services), th mud logger, by continuously monitoring mud gas content, shou ld be aware of situations of potential drilling hazard . It is therefo re usual fo r the mud logging unit to be equipped with a level monitor for the active mud pit. This allows the mud logger to be a second line of defense, after the driller, in detecting a well kick or loss of circulation. butane,
The Mud Log Format ’
Fig. 52.8 sho ws a typical modem mud log. There are currently no industry standards for mud logs, and presentation varies among operators. However, the track order commonly follows that shown in the example. Truck I is used for rate of penetration (RO P). Also included in this column are items of drilling data that may affect interpretation of the log (bit types, changes in drilling parameters, circulation breaks, etc.). Track 2 is for depth notation and for symbolic representation of special evaluations (for exam ple, c ored or tested intervals). Truck 3 is a graphical representation of formations penetrated . Usually the column is subdivided into 10 equal columns and graphic symbols are used to represent 10% increments of lithology types seen in cuttings. Unlike other track s on the log, the lithology track is not a calibrated p hysical m easurement but a subjective assessment. Care shou ld be taken to establish rules of drafting cceptable to the preparer and user of the log. Even after removal o f cavings and contaminants, cuttings sample is not truly represen tative of a single depth interval. Variation in particle size and density cause differences in annular recovery rate and mixing of cuttings in the annulus. A true cuttings per centage will never sho w sharp formation boundaries as a result of this mixing. For examp le, a thin sandstone w ithin a massive
PETROLEUM
52-12
IOLE
XL MUD LOGGING COMPANY COMPANY
ABC OIL COMPANY NETHERLANDS
WELL
DESMOND
F,ELD
ANDORRA
REG,ON
OF THE
5’ 02
‘0’ 50” 1530’E
DATE
CONTRACTOR
SCALE
A,
5340’
AT
SEAWATER GEL
TV
2300’
KCL POLYMER
TO
5345’
SYMBOLS
5345 DEEPER DRILLING CO
CHARLIE JONES’JA CKUP
LOG INTERVAL DEPTH FROM 400 DATE
2995
\EBREVIATIONS
AKBlo MSL 84 RKB lo SF 174 2
TOTAL DEPTH ,q,G , TYPE
7
735 *T
.ITHOLOGY
514!78
ELEVATION
20’AT
4185
TO
API WELL INDEX NO SPUD
9”aAT
IUD TYPES
DUTCH NORTH SEA
COORDINATES
:ASING RECORD 30 AT 400 x
Xl
SIZE
FROM5478 -UNIT500
LOG PREPARED
TO TO
5345 20 5 78
69, STANDARD BY
A EVANS,G
JONES
f3 EDWARDS
FORMATION EVALUATION LOG
Fig. 52.8-Mud
log formal
ENGINEERING
HANDBOO
MUD LOGGING
shale with sharp boundaries shown clearly by ROP and total gas analysis may app ear fr om cuttings to be a sandy shale horizon of much greater vertical extent. A geologist may use all available data to prepa re an interpretive lithological log. That is, in the previous example, to sharply sh ow the sandstone boundaries as indicated by ROP. Sometimes a mud logger will attempt to Again, in the example the mud logger would show the presence of sand in all the cuttings samples but exaggerate the percent sand in the sample coinciding with th higher ROP. This “semi-interpretation” may result in confusion when later sample examinations are compared with the mud log and, in my opinion, mud loggers should be instructed to prepa re a true cuttings log representing the percen tages of lithologies actually seen in the sample . If the mud logger is geologically qualified and the operator’s geologist requires geological interpretation, then an interpretive lithological log should be prepa red as an additional track on the mud log. Track 4 presents the results of hydrocarbon analyses. It may consist of one single width track but most often is subdivided into two or more separate tracks, as in the example. Track 4 will include the results of total gas, cuttings gas, and chromatographic analyses and when oil shows occur, an estimate of oil show quality and oil cut will be added . Supplemen tary gas analyses (helium, hydrogen, CO 2, or H 1 S) also may be added to this track or plotted on a supplementary log. Track 5 primarily is used for brief samp le description s. Also included in this track are mud test results, casing and cementation record s, hole deviations, carbide test results, and many other operating data used in interpretation of the mud log. O n wider format logs, Track 5 also may be subdivided to add an interpretive lithology and an extra data track to be used for the results of special analyses or calculations.
52-13
thin horizons may be submerged in a high background and not identified from total gas alone (Fig. 52.9). Even gas reliably identified as resulting from a drilled interval may be misleading as a show-qu ality indicator. Factors that affect the magnitude of a gas show include the volume of the rock cylinder crushed in the drilling proce ss, c ontrolled by bit diame ter an d RO P and dilution of liberated gas in mud ( i.e., the flow rate of drilling mud passing bottom as the hole is cut). Thus, gas show magnitude will be expe cted to increase in larger diameter holes, at higher RO P’s, or at reduced mud flow rate (Fig. 52.10). A simple technique is available to remo ve these facto rs from evaluation by normalizing gas show magnitude to standard or normal” set of drilling conditions:
(4 OB
where pOs = observed total gas, % G,, = normalized total gas, %, 4oB = observed mud flow rate, m”/s, “normal” mud flow rate, m’/s, 9n observed bit diame ter, m, OB d, = “normal” bit diameter , m, = observed ROP, m/s, and it = “normal” ROP , m/s. Once a
normal” set of conditions are selected, the equation can be readily simplified giving
0.010
G,, = Gpo~
OB
Interpretation
The object of logging drilling-mud gas shows is to identify potentially productive oil and gas horizons. While such zones often may be indicated by major events-e.g., large gas and fluorescence show s-mo re critical interpretation is required t o avoid false alarms or missed opportunities.
an
G,,
=
0.0126 Gpos90B (d,B)2
RoB
)
. .
.
. .
. .
. (5)
where Total Gas shows
The m agnitude of a total gas show is not in itself a conclusive indication of show quality. Gas detecte d at the surface origina tes in three ways: (1) from the disintegration of a cylinder of rock by the drill bit as the hole is deepened, (2) from the influx of gas from previously drilled formations expo sed in the borehole wall, and (3) from the drilling mud itself in the form of recycled oil and gas and decomposition of mud additives. In extreme cases (for example, in long, geopressured shale sections or when using oil-based drilling fluids) influx or contamination may constitute th e majority of gas seen at the surface. In such circumstances the magnitude of a gas show from a potential reservoir must be evaluated against the established backgrou nd gas level from overlying sediments. Gas show s from relatively
9n = 0.050 m3/s (793 galimin), d, = 0.251 m (9.875 in.), and R, = 0.010 m /s (118 ft/hr). Norm alization can be very useful in correlation of gas shows between wells drilled with very different programs. However, it should be remembered that normalization cannot remo ve the effect of influx and contamination; nor does it account for varying gas trap efficiency with ambient conditions. Finally, remem ber that the gas produ ced by drilling is liberated by the crushing of material at the bottom of the hole and is representative of the fluid composition within the rock pore space at the time of impact. R emember that oil and gas flow from a producing well; they are not mined. The presenc e of a fluid within a rock is not
52-14
PETROLEUM
Fig. 52.9-Mud
Fig. 52.10-Variation
of total gas with drilling parameters.
necessarily indicative of productivity of that fluid ,fram that rock. Comparison of total gas analyses from mud can yield useful clues and cuttings and chrom atograp hy as to the productivity of a hydrocarbon-containing formation. Total g as from the cuttings blender test is a good inare
carried
to
surface
and
relieved
of
formation
ENGINEERING
HANDBOOK
log total gas shows.
hydrostatic pressure, exsolution and expansion of gas should effectively flush the cuttings, leaving only a small volume of residual fluid at the surface. When h igher cuttings total gas, relative to ditch total gas, is observed, this is an indication that this flushing has been im peded. An obvious explanation is that the rock lacks sufficient permeability to allow gas expansion and flushing. How ever, residual low-gravity oil or tar will have a similar effect in impeding gas exsolution and escape . Inspection of cuttings lithology, fluorescence, and the cut test should provide confirming evidence (Fig. 52.1 I). For exam ple, strong cementation or shaliness in the cuttings would be indicative of low permeability, wher eas dark or dull oil stain and fluoresce nce with a slow cut or absence of cut is more indicative of heavy oil. At the opposite extreme of mobility, a formation may be so permeab le that it is flushed effectively with mud filtrate even before being drilled. On recovery to surface neither mud nor cuttings w ill contain hydro carbons. Indeed, no cuttings may be seen since such formations are commo nly unconsolidated and disintegrate on recovery. the first observations will be In this circumstance sharp increase in rate of penetration followed , after the lag time, by a negative” gas show; total gas declines below the original backgrou nd. Testing the desander effluent or a mud sampling probably will show an increase in loose sand grains. The negative g as show s confirms only the excellent permeability, and for this reason alone the zone deserves closer inspection when a resistivity log is available. No evidence is available of the formation’s fluid content from the mud log. Mo st potential reservoirs fall between these two extremes: producing (1) a positive gas show and (2) cuttings blender gas, depending on permeability and oil
52-l 5
MUD LOGGING
ss. LT GRI-B”FF. sue *NG. GO POR. S-iAlN. GO PL IEL ~LT,~L~u~H CUT
PL AS:OC C
ss.
F-NED
FRI.
SL TST. OIL GOL D
SD”
Y”.
SRT.
FY, N “IS F LOR FLOR,
GR.
PR
W/lNT0W
ABU W
FLOR,
BR
STRMG
CU T
IEL YEL-
FL OR
S. BRN, NEO-CRS, W/RN0 OTL. W SRT. BRN OIL S TN . EVE N ,....
!:-I. ;:!:{.‘.: ..I.. ,/. . , .:,. :. ... :... ‘. “::” ‘I ::: I:..:: I ;:: : --f-e++ ,:I; 1;: i : : i , *. ,. .::. ::._.. :;I,:;’ j .: :, :.,: r:;.:l: : ,./. p-’ ..&
0 ”LL YE0 “EL
OL
*EL -BR N FL OR, STRAW CUT. BRT CU T FL OR
5%
LT
GR. FRI.
SUB WI
GRY-ERW.
YE0
RNO. W SRT. S1L CUT. DULL
ORN G BRN FL OR. BRT L T YEL CU T
SL O FL OR
PETROLEUM
52-16
gravity. Color and fluorescence of the oil stain and results o f the cut test are indicative of oil type. How ever, note again that presence of hydrocarb ons in the rock and even presence of porosity and permeability are not conclusive evidence of hydrocarbo n productivity. Chromatogram
Interpretation
The hydrocarbon chromatogram is often a useful guide to reservoir productivity. It is not the actual am ount of any one alkane that is significant but the relative am ounts of light and heavy compon ents characteristic of the overall reservoir fluid. Such characteristics normally can be recognized on the mud log itself. An aid to this can be the calculation and plotting of the numerical ratios of the values of the various hydrocarb ons (e.g., C2/C I) C3/C 1, etc.). Such gas-ratio p lots will often yield or “events” not always imdistinctive “characte r” mediately evident from the chromatogram itself. However, interpretation of the plots depends on the same logical procedures. Most petroleum hydrocarbons originate from a similar organic source and proce ed in their maturation by similar tempera ture- and pressure-controlled physicochemical p rocess. F or this reason, petroleum accumulations, although m arkedly different in compo sition, tend to show a spectral relationship to each other in terms of the type and amount of hydrocarbon species present. Therefore, although two crude oils of 30”API gravity may be extremely different in total composition, they will contain so me similar comp onents in similar compositional relationships to each other. Since petroleum maturation continues by the continuous “cracking” of complex branch molecules into simpler straight chain molecules, these significant relationships are readily seen in the petroleum gases methane through butane. Thus, study of chromatographic analysis of these gases often may lead to a gross estimate of the type and quality of the reservoir. At low tempe ratures and shallow burial, biological and catalytic decom position of organic debris results in low-yield production of methane and CO2 Though the CO2 will dissolve in and migrate with pore waters, the methane will accumu late in porou s zones either as free gas or in solution in water. Such zone s will yield impressive gas shows when drilled, but with few exceptions will produce only gas-cut water. It is a reasonable general rule that a gas show containing methane as the only significant componen t is not comm ercially productive and is unworthy of further evaluation. How ever, it also should be remem bered that exceptional cases do occur, including the southern Nor th Sea nonassociated gas reservoirs that produce +94% methane. At higher temperatures, organic material first polymerizes to form kerogen, which is then hydrogenated and cracked with increasing temperature to form bitumens, tars, and progressively higher-gravity oil and gas. Associated petroleum gases are fragments of this cracking p rocess and as cracking continues, the proportion of light to heavy gase s increases in a manner similar to the lightening of the liquid hydrocarb on. This fractionation of gases and liquids continues during the migration of the hydrocarbons from source to reservoir
ENGINEERING
HANDBOOK
The end result of this process is reflected in a gas show chrom atogram . A gas-produ ctive interval will show predominantly methane and ethane with only traces of the heavier alkanes. Oil productivity is signaled by the enrichment of the heavier alkanes, especially p ropane. Decreasing oil gravity is reflected by progressively greater proportions of propan e and the butanes. In lowergravity oils, the concentration of these gases may exceed that of ethane. But all productive horizons will contain methane as the dominant alkane. Zones in which methane represents at least half of the total gas usually contain heavy residual oil from which the lighter gas and liquid fractions have migrated , leaving an immovable, nonproductive fluid. These general rules of chrom atograp h evaluation can prove useful in reservoir evaluation but of course sh ould not be used in isolation. Conclusions regarding oil gravity and mobility should be compared with the results of blender tests, cuttings examination, fluorescence, and cut tests. Furthermore, evaluation should proce ed from the prior-show baseline values and throughout the show interval. Considering the variables inherent in the drilling, transportation, and extractions of samples, no conclusion may be drawn from a single sam ple o r analysis. Conventional
Mu d Log
The conventional mud log offers mo re drilling and formation evaluation data in a single form than any other data source. Many of these data are subject to uncontrolled variables in the measurem ent technology and by the very nature of borehole environment. As a result, no simple conclusive rules of quantitative log evaluation are available. How ever, integration of all the data on the log with geological and regional ex perience can make the conventional mud log a most power ful e xploration tool.
Petroleum Engineering Services Geopressure
Evaluation
Th petroleum engineering functions of mud logging developed from the introduction of a number of pressure evaluation techniques during the 1960 ’s. These techniques are practiced conveniently in the logging unit and in some case s use data or equipment already available in the unit. Also, interpretation of the resultant data rcquires the same integrative appro ach, using drilling da ta and geological evaluation, as required in mud log interpretation. Unlike mud log evaluation, how ever, p ressure evaluation techniques are able to provide reliable quantitative estimates of fo mation parameters, such as pressures and porosity. lo Drilling into a geopr essured zone causes a change in number of basic formation/drilling relationships. This change is usually a reversal of a gradual d th-related trend in a lithologically uniform formation. IP Compaction increases uniformly with depth in a normal pressured clay rock. A geopressured zone may be poorly comp acted relative to those zones overlying it. Porosity and water content decrea se uniformly with depth in normal pressured clay rock. A geopressured zone i which dewatering has been slowed will show a reversal in the trend, w ith increased water content and increased porosity. Othe r factors relating to fluid movem ent, such as ionic concentrations, hydrocarb on saturations, etc .,
MUD LOGGING
52-17
may be different in geop ressured zones. D ifferential pressure across bottom is the difference between the drilling m ud hydrostatic pressure and the fluid pressure in pores of the undrilled formation at the bottom of the hole. Since drilling mud usually is denser than formation fluids, this difference will be positive and will increase with depth. In a geopressured zone, the formation pore pressure is abnormally high and the differential pressure across bottom will decline or even becom e neg ative. Thus, any measureab le param eter that reflects any or all of these factors may be used as a means of interpreting changes in formation pressure and eventually as means of evaluating and obtaining quantitative estimates of formation pore pressures.
Gas Analysis. The incursion of formation fluids into the borehole may result from a number of causes, some but not all of which result from an under ba anced condition-either tempo rary or permanent. ‘* If an underbalanced condition exists, th ere will be a natural tendency for fluid to flow from the formation into the borehole. With a formation having good porosity and permeability, this flow will be massive and a kick will occur. Such a kick will be indicated by the incursion of formation fluid downhole, causing the expulsion of mud from the boreho le at surface. Wer e this to continue, a blowout would result. It is the logging geologist’s responsibility (other duties perm itting) to monitor the mud pit level and to report any unpredicted or unexplained level changes. A massive incursion of fluid resulting in a well kick is unlikely to be misinterpreted as a gas show. In fact, if the hole is full, the kick should be recognized by a rise in pit level long bef ore the fluid causing it has time to appear at surface. How ever, minor incursions caused by slight or temporary underbalance, or whe re insufficient permeability to provide a sustained kick exists, do occur and must be interpreted correctly. When an underbalance sufficient to cause a kick exists but there is insufficient permeability to sustain a massive fluid influx, a steady fluid “feed-in” may result. If this minor flow is from a discrete formation already cut, it will be noticeable - producing a sustained minimum gas backgroun d even when circulating but not drilling. If this is the case, the logging geologist should m ake a note of this sustained circulating gas on the mud log.
If the feed-in is from the formation currently being drilled, then as a greater and greater ar ea of formation in the borehole wall is expo sed by drilling, increasing flow will take place. If this is the case, the mud gas will exhibit a sustained minimum when circulating but will consistently rise as drilling procee ds. Cuttings gas will inevitably be high relative to mud gas since it is only the lack of permeability that is preventing the feed-in fro becoming a kick. Where permeability is effectively absent (e.g., in clays or shales) even a minor feed-in cannot take place. Fluid pressure in the rock will gain access to the boreho le by the opening of pre-existent microfrac tures and partings in the rock. The result w ill be the caving or sloughing of rock fragments into the borehole, accompanied by a small amount o f gas. A minimum gas backgrou nd and, in this case, cavings recovery exist when circulating without drilling.
Fig. 52.12-Connection
gas indicati ng underbalance.
Circulating bottom hole pressure is higher than when the mud is static. This is caused by annular pressure losses when circulating. It is therefo re possible for feed-in, caving, or even a kick to result because of resultant underbalance when circulation is stoppe d. Furthermore, pressure is further reduced because of the swabbing effect when pipe is moved upward-e.g., when making a connection. The literal meaning of “swabbing” is the pulling of a full-gauge tool from the hole, acting like the plunger in a syringe and initiating fluid flow into the boreho le. Swabbing by moving the drillstring does not work in this way. When pipe is pulled upward, the high-viscosity gelled mud will attempt to move with the pipe, thus reducing the effective hydrostatic pressure acting on the borehole wall. Pressure reduction is a function of pulling speed, m ud rheology , and annular diameter . The important consideration is that pressure reduction takes place not just below the bit but at all points in the open hole. Downtime gas or connection gas is a gas show resulting from the mom entary underbalance caused by pump shutdown and/or pipe movement. It can be recognized by the occurrence of discrete gas show appearance at, or slightly less than, the lag time after circulation recom mences. This is gas actually being produced by the formation and, while not being plotted on the mud log, the value should be reported on the log because it is indicative of formation permeability and fluid content. When a connection gas occurs, the logging geologist also should check a flowline mud sample for evidence of produced oil or salt water with the gas. The incidence of connection gas should be reporte d to the
PETROLEUM
52-18
ENGINEERING
HANDBOOK
NORMALIZED
TOTAL GAS
:ONNEiC GA:SE
Fig. 52.13-Normalized
drilling supervisor, wh o may choo se to increase m ud density in response to the indicated underbalance. It is important to remember that the entire open hole section will be underbalanced by swabbing. The connection gas may not come from the bottom of the hole but from som e horizon above. In fact, two or even more connection g ases may result from a connection. For this reason it is important that lag time and annular velocities should be identified accurately by the logging geologist so that connection gases can be identified with the producing formation and the mud log annotated accordingly. Drilling into a permea ble reservoir with an underbalance is potentially dangerou s because a kick may result. Even if a kick does not occur im mediately, the hazard ous situation w ill be mar ked by an increasing feed-in as more formation is drilled, acco mpan ied by progressively larger connection gases (Fig. 52.12 ). The condition should be reporte d by the logging geologist and noted on the mud log. If increases in mud density alleviate or remov e the effect, this should also be noted on the mud log in explanation of the consequent reduction in gas. Fig. 52.13 demonstrates the effect of varying differential pressure on gas show magnitude. The total gas curves fo r two wells drilled thro ugh a similar section are shown. The data for both wells have been normalized to reduce the effects of hole diameter, ROP, mud pump output, and surface extraction efficiency. Well A was
total gas.
drilled with a constant mud density, whe reas in Well B mud density w as controlled to maintain a constant positive differential pressure (overbalance). In the upper portion of the section, the two gas curves are similar and the normalized gas curves coincide almost exa ctly. In the lower portion a progressiv e deviation between the two wells is seen that is somewhat reduce d but remains evident even in the normalized curves. We can interpret this as being caus ed by the penetration of a transition zone into a geop ressured zone. In Well A, maintaining a constant mud density results in a decreasing overbalance and eventually an underbalance or increasing negative differential pressure Connection gases occur and become larger with deeper penetration. Additionally, feed-in of gas from the underbalanced boreho le wall causes an increase in backgrou nd gas that, since it is not a product of freshly cut formation, cannot be accounted for in the normalization calculation. Well B, on which a constant overbalance was maintained by increases in mud density, did not show increases in gas backgrou nd or connection gases. Indeed, if any zone showed g ood permeability, the overbalance may have resulted in flushing gas away from the borehole and a reduction in observed total gas. By careful observation of these phenomena, a fairly accurate log of differential pressure (and hence pore pressure) may be obtained. This information should be used in conjunction with the other techniques described in the following paragraphs.
52-19
MUD LOGGING
Cuttings Evaluation. During the normal mud-logging process, cuttings are sieved and graded to a size assumed to be representative of drilled cuttings. The larger fragments are cavings from the walls of the borehole and play no part in the compilation of a lithological log. In geopr essure evaluation, these cavings play a major role. The presence of cavings in the sample indicates that the boreho le w all is unstable. The most noticeable an usually m ost predominant cavings ar e those of clay, shale, or calcareou s lithologies. Coal, howev er, will cave as a matter of course, hence interpretations should not include co al cavings. The amount of cavings in the bulk sa mple is an indication of the degre e of instabilit of the borehole walls. S imply w atching the cuttings traverse the shaker screens will give a reasonable indication of the amount and size of the cavings in relation to the bulk sample. Cavings are prod uced by underbalanced drilling and stress relief. Ab rasion of the walls by the drillpipe will also cause cavings, but generally these will not be discernible from cuttings because of their small size. If the pore pressure is higher than the hydrostatic pressure in the borehole, the hydrostatic pressure differential will cause the pore fluids to move toward the boreho le. In an imperm eable formation, the resultant pressure gradient adjacent to the borehole wall may become great enough to overcome the tensile strength of the rock. When this occurs, the rock fails in tension and cavings are formed. All parts of the earth’s c rust contain stresses th at change w ith depth , area, lithology, history, e tc. Drilling a hole in the ground relieves some stresses other than those in the vertical plane, and the hole geom etry in relation to som e stresses acts to concentrate them. If the boreho le w all is not suppo rted sufficiently by the mud column, it may fail either (1) in comp ression from the vertical stress or (2) in tension from the horizontal stress, or both. The drilling process causes the formation of microcracks and fractures, and these act as areas of stress concentration and potential initial failure points. T hus i is sometimes noticed that part of a borehole may cave copiously for a short time and then becom e stable. T his is because of the removal of the damaged zone (i.e., cavings) adjacent to the bore/form ation interface. Formation is expo sed th at is more c oherent and lacks concentrations of stress, thus it absorbs the extra energy withou t failing. Cavings prod uced by underbalanced drilling are typically long, splintery, con cave, and delicate (Fig. 52.14a). Cavings produced by stress relief tend to be mor e blocky and can vary in size tremendously, depending on the formation character istics. Exam ples a re shown in Fig. 52.14b. Remember that if the cavings are clays, they may react with the mud and lose their distinctive mor phology . Interpretations based on reactive clays shou ld be pursued with caution. The quantity and nature of cavings should be regularly reported on the mud log or on a supplementary data log if pressure evaluation services are being performed. Shale Bulk Density. Shale density determination can be of great value since it provides information on the compaction of the shale. Under normal conditions, shale
FRONT
‘ACE
Fig. 52.14-Cavings
L1 PLAN
resulting from underbalance
relief.
and stress
density should increase with depth. Any deviation from this consistent trend may indicate that geopr essures exist. The magnitude of the bulk density ch ange will vary with the type and magnitude of the geopressure. Often, the bulk density w ill decrease, but in other ca ses it may remain c onstant or continue to increase but at a lower rate than the previously established trend. S everal methods are used for measurement of shale bulk density. By using a container with Pycnometer Method. repeatable volum e, this meth od involves measuring change o f weigh t resulting from displacement of fluid by the sample. The most practical application of this method at the wellsite is to use a mud balance. Place enough cuttings in the cup so that the balance indicates 8.34 lbm/gal (i.e., density of fresh wate r) with the cap on. Fill the cup with water and weigh again. The new reading is W2 in the following equation. 8.34 -Ys= 16.68-w2
,
. ...
where ys is the specific gravity of sample and W z is the “mud weight” of sample and water, lbmigal. Mercury Pump Method. The bulk volume of a known weight of sample is measured. The bulk weight of a prep ared sample is first established using an accurate chemica l balance. The bulk volume of selected cuttings is then determined using a high-pre ssure mercury pump by the Kobe system (Boyle’s law principle) at a pressure of about 24 psi, which is recorded on the attached pressure gauge. Mercury is used to compress the air material. The high accuracy of the instrument and large amount of sample used (approximately 25 g = 2,000 individua shale cuttings) g ive good consistency of results. Be cause of the accuracy and convenience in operation, this method should be used whenever possible; however, very careful and consistent sample hand ling is necessary for best results. Buoyancy Method. The sam ple is weig hed in air and in a liquid of known density. This is an alternative version of the pycnometer method. Theoretically, it should be more accurate m ethod if an accurate laboratory balanc
52-20
PETROLEUM
DATA
HANDBOOK
trap air and water and resu lt in low apparent densities. In addition, the fluids used have unpleasant odor s and some of them may be hazardous to health. Toxicity label should be checked for specific mixtures but it is a good general rule to use these fluids in a fume hood or with a vapor extraction system. It is comm only observed that shale density may decrease as much as 0.5 g/cm3 or more. If this reduction occurs ov er a significant depth interval, the calculated overburden gradient may reverse. The low-density zone also may change in lithologic charac ter. Fissility, plasticity, carb onate content, co lor change, and other differences may not be apparent. Measurements from cuttings from water-based muds usually are too low, simply because o f the adsorption characteristics of clays. Likewise, measurements taken from wireline logs can also give false indications. Specifically, the formation density log can be affecte d by a rugose hole and the shallow de pth of investigation may not read beyond the hydra ted zone. T he result is erroneously low readings, causing excessive calcu lated p orosities. The sonic log will be affected greatly by hydra ted clay s, resulting in very high transit times, high porosities, and too low densities. Values may be successfully obtained from these logs when water-based muds are used, but caution should be exercised as errors may exist, as explained earlier.
is used. In practice, it is most inaccurate since the density of the liquid w ill vary with ambient temp erature. Density Comparison Methods. The simplest of these is the “float-and-sink” meth od. Shale cuttings are immerse d in fluid mixtures of different densities in which they will either float or sink, depending on relative densities. This method is inexpensive and quick but is limited in sensitivity because of large difference in the densities of available fluids (approxim ately 0.1 to 0.05 g/cm 3 and easy contamination of calibrated fluids. Density Gradient Method. This co nsists of a fluid column in which density varies uniformly with depth. This is prepa red by the partial mixing of a light fluid (neothene) and a heavy fluid (tetrabrom oethane) in which beads of known density are suspended. A calibration curve of density vs. depth is prepared. Shale cuttings imm ersed in the column will sink to the level at which the ir density is the same as the fluid. De pth is recorded and density read off from the calibration curve. Both of the heavy liquid methods (density comparison and density gradients), while being quick and simple, have the disadvantag e of determining the density of individual cuttings. Special care must be taken to ensure that cuttings ar e true bottomhole cuttings, and several determinations should be mad e for each interval to avoid anomalous results. Six or eight cuttings sho uld be chosen that are representative and free of dust or cracks that may
SHALE
ENGINEERING
PRESSURE
Fig. 52.15-Shale
data log.
LOG
MUD
LOGGING
The best densities are those obtained from wells drilled with less reactive muds, such as oil- or potash-based fluids. Both actual cutting densities and log densities should be accurate because the clay remains in its virgin state. Increases in density bey ond the normal trend becau se of decre ased po rosity or calcification should be noted carefully since these may constitute caprocks above geop ressures. Precipitation of pyrite or high iron concentration results in abnormally high bulk densities in clays and shales. It has been proposed that in some wells the occurren ce of pyrite in shales maske d the density re duction caused by porosity increase. Careful m icroscopic examination of clays may indicate the occurren ce of very fine pyrite, and high iron concentration is indicated by a red/brown color cast. Pore pressure interpretations cannot be accom plished by using shale density if heavy minerals ar e present; how ever, since shale density is mainly used for qualitative purpo ses in geopressure evaluation, the role of the other geopre ssure indicators remains unchanged. Any decrease in density (withou t chang e in clay character) may be recognized as a pressure transition zone. Recognition of a normal bulk density trend line may be difficult because of degre e of scatter in the rectangular coordinate plot. A semilog plot considerably reduces this scatter, but the normal bulk density range (approximately 1.6 to 2.7 g/cm 3) results in a more distorted trend line and difficulty in recognizing deviations (Fig. 52.15). reactions take place beShale Factor. Ion-exchange tween an adsorbent solid and a solution. Ions bound to the solid surface are released into the solution and other ions from the solution becom e fixed at the surface. Ion exchange can proceed by the exchange of positive ions (cations) or negative ions (anions) but not both. T he reactivity of a solid compound in ion exchan ge reactions is governed by its specific surface (surface area per unit volume) and by the surface density of ion exchange sites (points on the surface where ions may be bound). Reactivity is expre ssed as cation or anion exchange capacity (CE C or AEC ) using units of milliequivalents (of suitable ion) per hundred grams (of the compound). Various clay types have different CEC’s and consequently different adsorption capacities. A smectite-rich clay will undergo diagenesis to illite with increasing temperature and ionic exchange. For diagenesis to proceed, water must be flushed from the clays. If potassium exchan ge cations ar e not available, a montmorillonite clay will lose its water but will not convert to illite. Thus, if this type of clay is drilled with a water-b ased mud, th e clay will immed iately rehyd rate and cause severe drilling problems. Shale factor is a measure of clay CEC . CEC will as clays convert from montmo rillonite-rich to illite-rich with tempe rature (and thus with depth ). Pure montmorillonite clays have a CEC of approximately 100 meq /lOO g. Pure illites show no swelling character istics, but their CEC is generally between 10 and 40 meqilO0 g. Kaolinites have a CEC of approximately 10 meq/lOO g. Of the most common clay types, it is only the smectite group (including montmo rillonite) that has an affinity for wate r. Thus, any clay zone that contains montmorillonite decrease
52-21
Fig. 52.16-Shale
factor response.
will have an affinity for water in an amount proportion al to the montmo rillonite content, and this will be shown by a proportional value of shale factor. Note that the shal factor as measured at the wellsite will not give values corresponding to actual chemical C EC. This is because of impurities in the sample, metho dology , experimen tal error, and the fact that the methylene blue dye (used in the titration) is a very large molecule and thus cannot be adsorb ed in interlayer sites. If the clay is calcareous, and calcimetty is also being performed, then the shale factor may be corrected for carbonate content as given by
Fstu =
10 100 - Cca*
xFsha,
. . . . . .
. . . . .
.
.
where $ht = true shale factor, meq/lOO g, sha = apparent shale factor, meq/lOO g, and cart, = carbonate con tent, %. For example, a calcareous clay has a carbonate conten of 37 % , and an apparent shale factor of 16: 10 Fsht = -1oo-37(16)=25
meq/lOO g.
Theoretically, shale factor should indicate whether montmo rillonite dehydration or comp action disequilibri um was the major mechanism in generating an apparent geopressure. Geopressures caused by compaction disequilibrium indicate th at the pressur ed zone is imma ture with respect to shallower, normally pressured sediments. This implies tha t diagenesis has been restricted by the inefficiency o f the dewarer ing m echanism, resulting in clays containing a larger proportion of montmorillonite within the geopressured zone. Shale factor would thus decrease to the top of the geopressured zone, increas within the zone, and then decrease as the pore pressure gradients decline (Fig. 52.16). Any overall increase in shale factor within a geopressured zone indicates that compa ction disequilibrium has played a part in its formation.
52-22
PETROLEUM
EAR TH’S
-
HEATFLOW
LINES
SUR FACE
---
EQUITEMPE RATURE
Fig. 52.17-Distor tion of heat flow geopressured zone.
around
LlNES
an
insulati ng
If, however, a geopressured zone was caused by mont morillonite dehydration , then upon entering the interval a sharp decrease in montmorillonite content will be observed . Hence the geop ressured zone will contain less montmorillonite, because it has been converted to illite, which releases to the pore spaces water that has been unable to escap e fa st enough and results in a pore pressure increase. Shale factor thus will decrease in the pressured zone (Fig. 52.16). Shale factor cannot be a geopressure indicator. The differing responses described are not definitive, and geopressure has to be indicated from other sources before an interpretation by use of shale factor can be achieved. Geoprcssures caused by montmorillonit dehydration and comp action disequilibrium may cause no change in shale factor; also, if geopr essures were caused by another process (e.g., aquathermal pressuring that results when trapped pore fluids are heated but are unable to expand and is therefo re independent of matrix comp osition), a change may not be reflected in shale factor with depth. In the past, the consensus was that shale factor should increase in geop ressured zones and could thus act as an indicator. Re-evaluation of the various geopres sure mechan isms show that this is not necessarily the case. However, shale factor should be capable of delineating between compaction disequilibrium and montmorillonite dehydration as the major geopressure mechanism. Flowline Temperature. The geothermal gradient, the rate at which subsurface temperature increases with
gG
‘2-11
D -D
(loo),
.
.. . .....
.
2
where gG = geothermal gradient, “C/100 m, T, = temperature, “C (at depth D, , m), and ). T2 = temperature, “C (at depth 2, For any given are a, the geotherm al gradient is usually assumed constant. While the average gradient across
ENGINEERING
HANDBOOK
normally pressured formations may be constant, geopressured formation exhibit abnormally high geothermal gradients. l3 Since a constant flow of heat occurs radially from the earth’s core to the surface, the total flow of heat across any depth increment will be constant. How ever, the temp erature differential across an increment dep ends on the thermal conductivity of the material. Since overall heat flow from the earth’s surface is generally constant within any particular area, the heat flux throug h the various formations with depth is in equilibrium. The rate of change of temperature across a formation with a low therma l conductivity (caused mainly by high porosity) will be high; conversely, a low geotherm al gradient is indicative of high therma l conductivity formations-i.e., lower porosity. Water and hydrocarbon migration to shallower depths may also affect the geothermal gradient. Pore fluids, as insulators, retain heat so that on migration these hot fluids m odify the temperatures of the formations that they pass through and ultimately become trapped. Fowler14 cited examples from the Middle East, Canada, and U.S. oil fields of geothermal gradient bulges that indicated p ossible entrapment of hot fluids from greater depths. Th e mechanism also may be related dehydration, because the huge to montmorillonite volume of water squeezed from the clay provides the impetus for migration. “Dead ” basins (i.e., no source rocks) have been shown to exhibit normal geothermal gradients, hence on initial exploration wells the geothe rmal gradient m ay indicate the potential o f the whole area. An insulating zone produces a distortion in the isothermal lines that normally run perpendicular to the lines of heat flow (Fig. 52.17; Ref. 15 Because o f the high geothermal gradient, these are more closely spaced in the insulating zone. In the zones above and below, the isotherma l lines are more widely spaced in compensation and the zones exhibit a reduce geothermal gradient. The converse occurs in beds of high thermal conductivity (i.e., sand s and som limestones) Since water has a thermal conductivity of about onethird to one-sixth that of most rock matrix materials, it can be seen that therma l cond uctivity is directly related to the degree of compaction of a formation. The higherthan-normal water content of geopressured shales reduces the thermal conductivity. Therefore, the top of geopressured zone is marked by a sharp increase in geothermal gradient. Th temperature of the mud at the flowline may reflect the geotemperature, and recording of flowline temperature is a practical m ethod to determine temperature gradient, provided variable factors such as pump rate, lag time, am bient temperature, lithology, and temperature changes at the surface that ar caused by mud mixing and chemical treatments can be accounted for. In areas where large annual temp erature variations o ccur, considerable differences may be noted in flowline tempe ratures; even diurnal tempe rature fluctuations may cause a 10°C variation in flowline temperature while drilling. Prior to reaching a geopressured zone, a temperature transition zone will be encountered in which , because o distortion of the isothermal lines, ther e will be a reduction in geotherm al gradient. In practice, this effect is
UD LOGGING
52-23
TEMPERATURE
DATA
Fig. 52.18-Flowline
temperature
reflected in the flowline tem perature gradient, even to the extent of a fall in flowline tempe rature (i.e., negative gr adient), followe d by an extreme ly large increase in flowline temperature as the geopressured zone is penetrated (Fig. 52.1 8). A dual temperature probe system with sensors at the flowline and suction p it is effective in removing surface effect, if lagged differential temp erature is plotted. It is normally sufficient for the points to be plotted at 30-e intervals unless more frequent tem perature variation is noticed, but points plotted at 104 intervals allow more accurate d ata and better resolution for improv ed interpretation. C irculations, mud additions, water additions, and other significant events should be noted. It is found th at the resultant tem perature curve is broken when the bit is changed , or during short trips or other down time, and a certain time is necessary for the mud system to re-establish a tempe rature equilibrium upon circulation. The rate at which this thermal equilibrium is re-established may be significant. as mor e rapid re-establishment may indicate an increased geother mal gradient. Drilling variables that affect the
LOG
log.
rate of re-establishment of equilibrium include total mud volume. The practice o f reducing active pit volume to minimum, dictated by hole size, aids in reducing the reduce s the circulation time needed to stabilize flow line temp erature. A discontinuity in the plot also occurs at each casing depth and corresponds to a change in hole size. A higher annular velocity in open hole reduce s the amount of heat gained from exposed formations, and a lower annular velocity in the marine riser increases the amount of heat lost to the sea. However, these factors only lead to a change in measu red tempera ture; the rate of change of temperature should remain unchanged. Since pressure predictions can be based on temperature depth segment betwee n discontinuities can be analyzed separately for gradient trends. It is also helpful to replot a smoothed curve of segments end to end without regard for absolute tempe rature values. In certain cases it has been found th at, instead of plotting the individual segmen ts as an end-to-end smo othed cu rve, end-to-end plotting o f the individual segment trend lines may be of
PETROLEUM
52-24
ENGINEERING
HANDBOOK
removing the need to derive emp irical matrix strength constants, but making d-exponent lithology-specific as in
d=
where = drilling exponent (dimensionless), ROP, ft/hr, N = rotary speed, rev/min, = WOB, lbm, and db = bit diameter , in. R
IIFFERENTIAL P RESSURE ,CROSS BOTTOM (PSI)
RATE O F P E N E T R * T ,O N (CONSTAN T N.W.0.EC0)
Fig. 52.19~-Response
0°C
of drilling rate to geopressure.
This trend-to-trend smoo thed curve is merely a graphical method of removing irrelevant scatter from the plot. The reduction in tempe rature gradient cau sed by distortion of isotherma l lines may be noticed b efore the geop ressured zone is encountered; that is, an advance warning of geopressure may be given. Thus a fall i flowline temp erature gradient fo llowed by a sharp rise when the geop ressure transition zone is drilled provides a warning that even closer attention must be paid to other drilling p aram eters to achieve confirmation of possible geopressures. However, like other methods of pressure evaluation, flowline temp erature reflects a varying physical parameter in an assumed constant rock type; in lithology must be closely therefore, change s monitored to avoid false indications. value.
Drilling Models. Bingham I6 proposed that the relation ship between ROP, weight on bit (WOB), rotary speed, and bit diameter may be expressed in the general form
-=
,.......................
(9
where
R = ROP, ft/min, N = rotary speed, rev/min, db = bit diame ter, ft, = WOB, lbm, Kn = matrix strength constant (dimensionless), an = formation “drillability” exponent (dimensionless).
Jorden and Shirley ” solved Eq. 9 ford, inserted constants to allow comm on oilfield units to be used and to produ ce values of d-exponent in a convenient work able range. Most important, however, they let KD be unity,
In a constant lithology, d-exponent will increase as the depth , comp action, and differential pr essure across bottom increase. Upon penetration of a geopressured zone, comp action and differential pressure will decrea se and be reflected by a decre ase in d-exponent (Fig. 52.1 9). Differential pressure is dependent on mud density as well as formation pore pressure. Therefore, any change in the mud density used promo tes an unwanted change in d-exponent. Rehm and McClendon” proposed a “mud weight correc ted” drilling exp onent of the form
dxc where = correc ted d-exponent (dimensionless), R ROP, ftlhr, N = rotary speed, revimin, dh = bit diamete r, in., g+, = normal formation balance gradient, lbm/gal, Pe = effective circulating density, lbm/gal, and = WOB, 1,000 lbm, d,
or in the metric form
.
(12)
in tonnes (1,000 kg), db in with in m/hr, N in rpm, cm, and g,p and ,oec in gicm3. This correction was empirically derived b ut has been applied worldwide with much success. The use of actual mud density in place of effective circulating density (ECD ) has been found to be acceptable within normal limits of accuracy. ECD should, however, be used when available. Factors no t considered by d-exponent in its basic form are drilling hydraulics, tooth efficiency, and matrix strength.
MUD LOGGING
52-25
0°C
SEMlLOG
SCALE
of formation pore pressure gradients for Fig. 52.20-Example d, plot.
Drilling hydraulics becom e important in large holes whe re efficient ho le cleaning is impossible and in soft formation whe re jetting w ill mak e a large contribution to drilling. Matrix stren gth controls both magnitude and rate of change of d-exponent with depth. Tooth efficiency can affect d-exponent in two ways: (1) tooth w ear will cause a gradual increase in exponent (i.e., decrease in ROP), and (2) a change of bi type may produ ce a change in d-exponen t, especially if the change is a radical one (e.g., from milled-tooth bit to an insert or diamond bit). If differential pressure beco mes large, the simple ratio correction to the d-exponent will not eradicate the effect on ROP. Furthermore, the relationships among force applied, Wldb, otary speed, N, differential pressure , g,,~,/p~(., and ROP, re more complex than the d-exponent formulation would imply. While working well within certain normal w orking ranges, radical chang es in any of these parameters (for example, change in hole size after setting casing) may result in a shift in d-exponent trend. When plotted on a logarithmic scale against depth, the d-exponent will exhibit an approxim ately linear increasing trend throug h normal,” hydrostatically pressur ed formations. W here geopressure, abnormally high formation fluid pressure is encountered, d-exponent values will fall consistently below the extrapolation of this normal trend. It has been show n empirically that d-exponent deviation may be related to formation pore pressure anomaly by the simple ratio -Ppa
Ppn
-
dxcn )
.xco
.
. .
.. .
..
.(13)
Fig. 52.21-Logging
where PPa
=
actual
unit systems.
pore pressure at depth of interest
lbmlgal equivalent m ud density (EMD ), PPn
= normal pore pressure, psi, or formation balance gradient, lbmlgal (EMD), observed corrected d-exponent at depth of interest, and expected corrected d-exponent on normal trend line at depth of interest.
Using this relationship, it is possible to calculate pore pressure or formation balance gradient (equivalent mud density to balance pore pressure ) from d-exponent. Alternatively, the relationship may be used to prepa re an overlay allowing direct reading of formation balance gradient from the d-exponent plot (Fig. 52.20). Drilling ex ponents may be calculated from driller’s data using a simple calculator and manual plotting and trend recognition. How ever, quality of the data is greatly improved when a data-acquisition system provides WO B, rotary speed, and mud density directly to the logging unit and a minicomp uter is used to read these senprint or plot the results. C omp uter-equipp ed
mud log-
52-26
PETROLEUM
ENGINEERING
HANDBOOK
it :, :::::t::::i::::l::1:i:X
Fig. 52.22-Formation
ging units were introduced to provide pressure evaluation services in areas of known geopressured problems (e.g., offshore U.S. gulf coast) and high-risk, hazardous exploration areas (e.g., the North Sea). Fig. 52.21 show s the available equipm ent configurations of com puterized logging unit. In addition to pore pressure, this type of unit commonly provides a pressure log including supplementary calculations of fracture pressure, overburden pressure, and kick tolerance (F ig. 52.22 ). Petrophysical Measurements Mud Log Data. As discussed previously, mud log data are qualitative in nature. It is not possible with conventional mud log measurem ents to obtain quantitative values of such param eters as porosity, permeability, hydrocarbon saturations, etc. Howev er, the mud logging unit may provide special equipment or services, allowing more quantitative evaluations. For example, while a mud log gas analysis cannot be truly representative of gas production comp osition, gas analysis of recove red fluids from a well test can be. Using conventional gas analyzers and chromatograph, a quantitative analysis of recove red natural ga s may be obtained. For more complex fluids, special chromatographs, pyrolyzers, or analyzers can give complete analysis of oils or sour gas. The logg ing unit also may provide ionic analysis of recovered waters and, by use of tritium ( 3H) or nitrate (NO3 ) tracers, provide discrimination between recovery of mud filtrate and true formation water. These types of service can be of special
pressure log.
value in remo te locations wh ere logistics o r distance prevent rapid transport of samples to an analytical laboratory. Core Analysis. In addition o gas and fluid analysis, conventional core analyses I9 (porosity, bulk density, permeability, and saturations) also may be perfo rmed in the logging unit when th e drilling operation is remo te from the laboratory. Wellsite core analysis offers the advantages o f rapid evaluation and high-quality samples from fresh core but is rarely of the high quality to be expected from the specialized equipment and personnel in core laboratory. A new technique, which is suitable to the logging unit, uses a pulsed nuclear magnetic resonance analyzer to determine fluid content, total and free-fluid porosity, and permeability. This device, w orking on a principle analogous to the nuclear magnetic logging tool (NML), provides accurate, repeatable data from minimal quantities of sample and without com plex sample preparation. Samples m ay be obtained without causing core or sidewall core destruction and the test may be performed on cuttings. Drilling Porosity
The d-exponent (Fig. 52.20 ) develops a consistent trend with depth controlled by increasing overburden loading and compaction. Changes in formation pore pressure gradient will result in major, con sistent deviations from this trend. The d-exponent data also will exhibit m inor,
52-27
MUD LOGGING
POROSITY
PERMEABILITY
ROCK
PROPERTIE
[MILLIDARCIES]
Fig. 52.23-Drilling
inconsistent scatter about the prevailing trend, reflecting continuous variation in rock mineralogy, cohesion, and porosity. Mo re sophisticated , second -generation drilling exponents are able to isolate the major pore pressure and minor rock chara cter variations. With this type of analysis it is possible to provide a continuous log of pore pressure and “drilling porosity.” to It is important remem ber that drilling po rosity, although s caled i percenta ge units, is not a true porosity measurem ent. It is primarily a rock strength indicator, reflecting b oth porosity and intergrain cohesion. As such, its response is very similar to that of the sonic log, and the two logs correlate extremely well (Fig. 52.23). Unlike the d-exponent, the second generation drilling exponents require comp lex manipulations and iterations, limiting their use to logging units equipped with a computer. Also, unlike the d-exponent, they do not involve widely published and used single method. Although based on similar drilling response models, all mud logging contractors offer drilling porosity logs involving their own unique mathematical methods that are commonly he ld as proprietary secrets. While understandable from a comm ercial view, this policy places th e user i the position of being able to judge the value and reliability of a particular log only on the bases of his or her own experience and limited pu blished results. It is hope d that, with the maturation of this type of service, wider publication and discussion of meth ods will begin.
FORMATION DENSITY
1~ PSEUD(
-SON1
porosity log.
Dr illing
Engineer ing
Ser vices
The mud logging unit can provide two levels of service of value to the drilling engineer-d ata acquisition and data analysis. Data Acquisition
An automatic data acquisition system loc ated in the logging unit will monitor sensors installed on equipment, flowline, mud pits, pum ps, etc. Simple calculations are perform ed on the data (e.g., calculation of total depth and ROP , summation of pit volumes, comparison of current values with high- and low-alarm setpoints). Results then are displayed on TV monitors at various locations around the rig and may be recorded on a printer or magnetic tape. By use of a dedicate d land line or satellite link, data can be transmitted to a remo te location, allowing several rigs to be monitored from a single central control room. This type of equipment was introduced by mud logging service com panies as a means o f obtaining drilling and mud data more reliably and rapidly than could be expected from standard rig instrumentation. While these data were required initially for pressure evaluation analyses, the data acquisition system provided an important secondary function as a rig monitoring service by supplying th e drilling engineer accurate, up-to-date drilling information while aw ay from the rig floor and a complete foot-by-foo t record of drilling prog ress and performance on paper or magnetic tape.
PETROLEUM
52-28
Since that time. sev eral conventional rig instrumentation manu facturers have upgra ded their product lines to include similar data acquisition systems that opera te i an unmanned, or “stand-alone,” mod e. While this offers the opera tor th e advantage of flexibility in selecting services (i.e., the mud logging service be st suited to the geologist and the data acquisition service best suited to the engineer), it does have drawbacks. Reliability of a data acquisition system is primarily controlled by the operation of its sensors. In the rigorous environment of the rig this requires regu lar attention. The success of a stand-alone data acquisition system is related entirely to the training and motivation of the ri crew or the availability of manufac turer’s service personnel. The mud logging unit is manned at all times. Trained personnel are available at all times to calibrate, maintain, and service the data acquisition system and its sensors. Since these personnel are already at the wellsite as part of the mud logging service, this extra margin of reliability is achieved withou t extra cost beyond the similar co st of the data acquisition hardware.
Data
Analysis
Beyond simp le data acquisition, the mud logging serv ice also may supply compu ters, software, and specialized wellsite personnel for drilling d ata analysis. The desirability of such services d epends on the difficulty and cost of the drilling operation, availability of oil company expertise at the wellsite, and quality of comm unication with the exploration headquarters. For exam ple, infill drilling in an established dom estic field using a well-develo ped drilling progra m, e xperienced wellsite supervisors, and close comm unication with home office requires data acquisition only as means of monitoring optimal and safe adherence to the drilling progr am. On the other hand, on an offsho re wildcat, the availability of data analysis nd expertise at the wellsite can be very cost effective. *O An increase in drilling efficiency or a decre ase in downtim e sufficient to save a single d ay of rig time can, in these circumstanc es, produ ce sufficient savings to pay for data analysis services for the whole well. Data analysis services offe red include: (1) bit optimization-selection of bit type and operating param eters to optimize bit RO P and bit life; (2) bit econom ics-cost per unit depth and breakeven calculation between bit types; (3) drilling hydraulics2’ -optimization of drillstring, nozzle, an d annulus hyd raulics; tomho le position, and intersection points for deviated wells; (5) trip monitoring-calculation of string weigh ts, swab pressures and fillup requirements for tripping, monitoring of pit level deviations, and overpull (frictional drag in the borehole); (6) casing calculations-assembly of casing tally, calculation of cement volumes and mixing requiremen ts, and monitoring of displacement; (7) pressure control-calculation of mud weight, olume and pressure requirements for safe well control lo; and (8) logistics-usage and inventory control ing, well progress data base, and report generation.
Selecting a
E N G I N E E RI N G
ud Logging
HANDBOOK
S e r v ic e
standard service: (1) standard mud logging; (2) mud logging and data acquisition, and (3) mud logging and data analysis (including pressure evaluation). For the most basic level of mud logging, a single opera tor m ay tje responsible for 24-hour operation. More sophisticated services and data acquisition usually require two geolog ists working 12-hou r tours. Data analysis services require tw o people, a geologist and an engineer. on each tour. Each of these services may be augmented with extra pyroanalyzers, more powerful com puters or peripherals and specialist personnel at extra day rate as the drilling program demands. 2,10 At least two mud logging contractors now offer an additional, fourth level of service in which m ud logging and data analysis are combined with an MW D service. Very little of the information gather ed by a mud logging unit is not obtainable from som e other sourc e. For exam ple, stand-alone instrumentation can monitor gas, mud, and drilling param eters; rig crew s can catch samples; porosity is available from wireline log s; and oil company geologists and engineers may perform geological evaluation and drilling data analysis. Why then is mud logging such a widely used service? The advan tage of mud logging service is that all these data may be derived from a single source, the mud logging unit, lo cated at the wellsite and continuously manned w ith dedicate d, specially trained personnel. Therefore the data are obtained more reliably, more quickly, and usually more economically than from any other combination of sources. Reliability and speed therefore are the tests required in selecting a mud logging contra ctor. The equipment must be designed and maintained adequately to provide reliable and safe operation in the rigorous wellsite environment. The wellsite crew must be trained to opera te, maintain, and troublesho ot the equipment and to understand its output. The contractor must maintain a dequa te service personnel and inventory to allow rapid repair or changeo ut in the event of major malfunctions. The logging crew mu st be trained in geological and engineering theory, be experienc ed in practical drilling oper ations, and have a thorough knowledge of the geological section, drilling p rogram , and operational proce dures of the particular well and operator. Once a contractor is selected, economy becomes the prime consideration in choosing a level of service. In day-rate drilling, time and money may be directly equated. Any service that speeds well progress, reduces down time, or prom otes decision-making is potentially cost saver. Even on footag e drilling, personnel, communications, etc., are cost-generating factors which may be reduce d by improv ed drilling efficiency. On rank wildcat explo ration wells, the “bird in the hand” philosoph y may be desirable to obtain d ata at the earliest possible time as a hedg e against the risk of it being unavailable later. For exam ple, to obtain porosity measurem ents from the mud logging unit while drilling is an investment against later boreho le loss or dam age that may prevent later wireline logging.
MUD LOGGING
52-29
Quantification of cost saving is possible by using the same metho ds u sed to calculate drilling cost per foot. I its simplest form this is Cbe
+
Crr/
)
.
...
.
(14)
For cost effectiveness (i.e., fo r the additional service to save its own cost or more), overall well cost must be unchanged or reduced. T hus,
A(cd
D,)=(cd
A(CdX D,)= where Cd C, Cbp 11
= = = = =
+CF2
+.
~CFn)+(Cdrl
X(t,+tt+t,+t,+td)];D,
XD,)Io,
-[E(c,,), x[C(t,),
AC,, I
.
.cdrn)
. . . . .
.(15)
AC,,
-At,,],
+
[~(c,,>,
+Af,,
CdxD,=C(CF),, C(C,,),c(t),,
I.....16
where CF = individual fixed cost items or footag charges services, $, Cdr = individual day rate services, rentals, salaries, etc., $, t, = rotating time, days, t, = tripping time, days, time (reaming, conditioning, to = off-bottom well control, etc.), days, = evaluation time (logging, testing, coring, t, etc.), days, and td = downtime (breakdowns, weather, decision making, etc.), days. Using this formulation, it is possible to calculate the decre ase in one cost catego ry required to offset an increase in any other. For exam ple, consider the use of drilling optimization. Let us assume , conservatively, that regional statistics indicate that by upgrading a mud logging unit to include data acquisition equipment no overall ROP improvement is obtained but that a well can be comp leted using one less bit. Well cost as result of this is given by
+W C,,),
ACFbl z(cdr) n Ar,l C(AZ[)~
= = = =
’ . ’’’
--Art, I,
(17)
(20)
$l,ooo $6,000/D 12 hours=0.5 30 days
Offshore
day
ACF~I E(c&)n Atrj E(Afl),
= = = =
$1,000 $19,OlWD 12 houn=O.S 35 days
day
we obtain for onshore:
AC,, I
1,000+(6,000)(0.5) (30 -0.5)
=$135.59/D. The extra equipment will result in an overall c ost saving on the well so long as it does not increase the mud logging daily rate by more than $135/D. We obtain for offshore:
AC,, 5
1,000+(19,000)(0.5) (35 -0.5)
=$304.35/D. Using addition drilling rotating
these same figures, let us now assume that, in to saving one bit, an overall decrea se o f 5% in time is also achieved. If I, =21 days, saving in time=21 ~5% = 1.05 days, then for onshore: 1,000+[6,000
XT
where ACFbl = cost of one bit saved, dollars, AC,, = extra cost of mud logging, d ollars/D, (f,),? = total time on location, days At,, = time for one trip saved, day, and (C = well cost. dollars, xD,)’
(19)
If this is evaluated as true, that the day rate fo r the extra equipment is in fact less than the evaluated expression, then the service is cost effective on the particular well. In this case, substituting som e reasonable figures such as:
I
-AcFbll
+Ac,,l[W,),
....
1
Onshore
xD,)‘=[C (c~)n
(18)
xAt,,l
E(t,), -At,,]
an
(cd
.
$,
+CdR.. . .
-Acml +AC ,,
drilling cost, $/m, rig cost, $/D, cost of bits and expendables, tim e on location, days, total well depth, m.
For optimization of services and products this can be expanded to the form Cd=[(CFI
XD,)‘-(cd
(0.5+1.05)]
(30-0.5-1.05) =$362.04/D.
For offshore:
AC,, =
1,000+[19,000
(0.5+1.05)]
(35-0.5-1.05)
=$910.31/D.
PETROLEUM
52-30
gc
P/m = n qoB =
developm ent drilling, especially w here problem s such as geopressure or crooked holes occur.
Nomenclature = cost of bits and expendables, dollars carh = carbonate content, Cd = drilling cost, dollars/m individual day rate services, rentals, dr salaries, etc., dollars Cf = individual fixed cost items or footag charges services, dollars C, = rig cost, dollars/D d = formation “drillability” exponent (dimensionless) = bit diame ter, ft or in. = “normal” bit diameter (m) = observed bit diameter (m) d,,. corrected d-exponent (dimensionless) d,,.,,expected corrected d-exponent on normal trend line at depth of interest d , , , observed corrected d-exponent at depth of interest = apparent shale factor, meq/lOO g Fh = true shale factor, meq/lOO g = normalized total gas (%) = observed total gas (R) po Cbr
, =
For and Status of Services
The terms “ mud logging” covers a diverse range of services and qualities of service. It is regrettable that, in the U.S. esp ecially, the whole industry is accor ded a status reflecting its lowest level. Field employ ees of the higher quality and more reputable contractors commonly eschew the term “mud logger,” preferring the title logging geologist” or logging engineer,” depending on their educational background. The wide range of equipment and techniques used by such companies commonly results in their personnel being the best educated and trained service personnel present on any wellsite. In 1980, the Sot. of Professional Well Log Analyst (SPWL A) established a Hydrocarbon Well Log Standards Committee comprising members of both the service company and exploration company sides of the field. The efforts of the committee have done much towa rd establishing standards and status representative of the best of the industy. 22,23 I express my gratitude to this committee and its members for these efforts and for assistance in the developm ent of this chapte r.
HANDBOOK
geothermal gradient, “C/100 gq% = normal formation balance gradient, lbm/gal D matrix strength constant (dimensionless) rotary-speed, rev/min Ppu = actual pore pressure at depth of interest, psi, or formation balance grad ient, lbm/gal equivalent mud density
These cost justifications, or cost savings, refer only to th e x t r a cost of data acquisition above that of standard mud logging and include only those benefits resulting from drilling optimization. Othe r cost savings resulting from better rig and mud monitoring and well control m ay also be quantifiable from a study o f regional drilling statistics. A cost benefit analysis of this type is a worth while approac h to all aspects of drilling cost reduction. Cost saving resulting from advanced evaluation and monitoring commonly is appreciated in expensive offshore exploration. As the above examples show, such techniques may
Standards
E N G I N E ER I N G
S td
t,
fl
Ol),
= I,
t, t, v ;;
VP
varl
= w=
(CdXD,)’
=
Fbl
=
AC,, = At,, = Per =
@MD) normal po re pressure, psi, or formation balance gradient, lbm/gal (EMD) “normal” mud flow rate (m3/s) observed mud flow rate (m3/s) pump output, m3/s) ROP, ft/min “normal” ROP (m/s) observed ROP (m/s) downtime (breakdowns, weather, decision making, etc.), days evaluation time (logging, testing, corm ing, etc.), days lag time, seconds total time on location, days off-bottom time (reaming, conditioning, well control, etc.), days rotating time, days tripping time, days annular volume, m 3 hole capacity, m3/m pipe capacity and displacement, m3/m annular velocity, m/s WOB, Ibm well cost, dollars cost of one bit saved, dollars extra cost of mud logging, dollars/ time for one trip saved, day effective circulating density, Ibm/gal
References 1. “Field Geologists Training Guide.” Exploration Logging Inc., Sacramento, CA (Jan. 1979). 2. “Mud Logging: Principles and Interpretation,” Exploration Log ging Inc., Sacramento, CA (Aug. 1979). 3. “Formation Evaluation-Part I: Geological Procedures.” Exploration Logging Inc., Sacramento, CA (Feb. 1981). 4. Hopkins, EA.: -“ Factors Affecting Cuttings Removal Rotary Drilling,” .I. Pet. Tech. (June 1967) 807-14; AIME, 240.
During Trans..
5. Sifferman, T.R. et al.: “Drill Cuttmg Transport m Full Scale Vertical Annuli,” J. Pet. Tech. (Nov. 1974) 1295-1302. 6. Low, J.W.: “Examination of Well Cuttings,” Quarterly cjj r/w (1951) 46, No. 4. l-48. Colorado School ofMines 7. Maher, J.C.: Guide book V/I/: Lqging Drill Curtings, Oklahoma Geological Survey, Norman (1959). 8. McNeil, R.P.: “Lithologic Analysis of Sedimentary Rocks.” Eu[(., AAPG (April 1959) 43, No. 4, 854-79. 9. Clementr, D.M., Demaison, G.J., and Daly, A.R.: Wellsite Geochemistry by Programmed Pyrolysis.” paper OTC 3410 presented at the 1979 Offshore Technology &if&-ence. Houston, April 30-May 3 10. “Theory and Evaluation of Formation Pressures: The Pressure Log Reference Manual.” Exploration Logging Inc., Sacramento, CA (Sept. 1981). I I. Hottman, C.E. and Johnson, R.K.: “Estimation of Formation Pressures from Log-Derived Shale Properties,” .I. Per. Tech. (June 1965) 717-22:
Trans..
AIME. 234.
