process control experiment

GAS FLOW (AF 922) PLANT CONTROL

1.0 •

• • •

SUMMARY

To To identify identify the important important component component of the air !o" contro# ytem and to mar$ them in the P%& 'iaram To To carry o*t o*t the tart+*p tart+*p proced*r proced*re e ytematica##y ytematica##y To To t*dy a ,o#*metric ,o#*metric !o" !o" rate mea*reme mea*rement nt *in the ori-ce p#ate p#ate To To t*dy a ma !o" rate mea*rement mea*rement *in the ori-ce p#ate and perfect a #a" to comp*te the a denity from the pre*re and temperat*re

• •

mea*rement To To t*dy a ma ma !o" rate (Fm) (Fm) contro# contro# *in P.' contro contro##er ##er To To dein a P.' P.' contro##er contro##er *in *in /ie#er Nicho# Nicho# c#oed #oop #oop method method

2.0 OBJECTIVES

.n the a !o" proce contro# ytem e0periment1 an e*ipment mode# AF 922 "i## 3e *ed There are i0 o34ecti,e need to achie,e "hich to identify the important component of the air !o" contro# ytem and to mar$ them in the P%& 'iaram1 to carry o*t the tart+*p proced*re ytematica##y1 to t*dy a ,o#*metric !o" rate mea*rement *in the ori-ce p#ate1 to t*dy a ma !o" rate mea*rement *in the ori-ce p#ate and perfect a #a" to comp*te the a denity from the pre*re and temperat*re mea*rement1 to t*dy a ma !o" rate (Fm) contro# *in P.' contro##er and to dein a P.' contro##er *in /ie#er Nicho# c#oed #oop method The e0periment "i## fo##o" the proced*re ytematica##y 3y fo##o" the #a3 man*a# in order to a,oid any error .n thi e0periment there are t"o indicator that *ed in thi e0periment "hich are Contro# of Ga F#o" Sytem and P.' Contro##er T*nin A## thee condition are needed to tet in order to tet it f*nction repone "hether fat or not .n the theory tate that 3y decreain the ,a#*e of P51 the oci##ation "i## increae and the repone dampin e0treme#y for the eco econd nd tria tria## at &678 &678 comp compar ared ed to 2778 2778 of et et poin point t At the the #at #at of &


e0periment1 there are oci##ation in the raft and P5 "a increae *nti# the oci##ation damped o*t "hich i a3o*t 9 8 3.0 INTRODUCTION

The air !o" proce contro# trainin ytem (:ode# AF922) *e air to im*#ate a a or ,apor phae !o" proce Ga;,apor ,o#*metric !o" can 3e mea*red either 3y the di
meter

Compenation for pre*re and temperat*re

,ariation *in the Perfect Ga La" can 3e made to the !o" mea*rement 3y the dia*en F1 27&7)

2


Fi*re &? .f A2;A& @ &; the ta3i#ity of the ytem i o$1 accordin to /ie#er and Nicho#

The /ie#er+Nicho#B c#oed #oop method i 3aed on e0periment e0ec*ted on an eta3#ihed contro# #oop (a rea# ytem or a im*#ated ytem) "hich i#*trated in Fi*re 2

Fi*re 2? The /ie#er+Nicho#B c#oed #oop method i e0ec*ted on an eta3#ihed contro# ytem An ori-ce p#ate 3aica##y a thin p#ate "ith a ho#e in the midd#e *ed to mea*re the rate of !*id !o" .t contrict the !o" and mea*rin the pre*re di

of the !*id and the ,e#ocity of the !*id When the ,e#ocity increae1 the pre*re decreae and ,ice ,era that preented in Fi*re  (eF*nda1 .nc1 27&&)

Fi*re ? Ori-ce P#ate 'iaram 4.0 RESULTS AND DISCUSSION

Start Dp Proced*re The re*#t "a ho"n in a raph form "ith three co#or of #ine "hich i red #ine1 the reen #ine and the 3#*e #ine The red #ine "a repreentin a the ma !o" rate1 the reen #ine repreent a the ,o#*metric !o"rate and 3#*e #ine for TE9& (C) At thi tae1 : of F.C9& "a et to +8 and it "a ad4*t the man*a# ,a#,e :977A ti## the air !o" rate i a3o*t 67 $;>r After that1 : "a et at &78 and chane to 678

Contro# of Ga F#o" Sytem A for tartin of the e0periment1 the P.' contro##er "a tart in tria# mode "ith a peci-c data inp*t that "a ho"n in Ta3#e & 5aed on thoe tria# mode1 the e*ipment condition can 3e identifyin Ta3#e &? The P.' Tria# a#*e 


P5&? 2778 T.&?  ec T'&? 7 ec

P52? &678 T.2?  ec T'? 7 ec

P5? &778 T.?  ec T'? 7 ec

Fi*re ? P.' tria#

5aed on Fi*re &1 the raph ho" that the red #ine "a the oci##ate condition The red #ine "a repreentin a the ma !o" rate1 the reen #ine repreent a the ,o#*metric !o"rate Accordin to the raph1 ma !o" rate in the red #ine i in a proper c*r,e "hich preented a3o*t 968 of ,a#*e that "a et in the e*ipment The e*ipment "a et in man*a# mode "ith "ith : H 268 for the -rt tria# and "ith : H 78 for the econd tria# Ne0t1 ma !o" rate in the red #ine in the econd tria# preented a3o*t 9I8 of ,a#*e A the re*#t1 the red #ine repond *ic$#y than the reen #ine and it tate that the ma !o" rate repreent more eniti,e than the ,o#*metric !o"rate 5y decreain the ,a#*e of P51 the oci##ation "i## increae The P5 correpondin to the contant amp#it*de oci##ation a P5J *#timate P5 i act*a##y the fractiona# error chane1 re#ati,e to the error rane1 re*ired to prod*ce a &778 chane in the proportiona# term o*tp*t The repone dampin e0treme#y for the econd tria# at &678 compared to 2778 of et point ettin that i##*trated in Fi*re  For thi e0periment1 the P.' contro##er *in /ie#er Nicho# c#oed #oop method "a *ed The point of t*nin a P.' #oop i to ad4*t ho" 6


arei,e#y the contro##er react to error 3et"een the mea*red proce ,aria3#e and deired etpoint .f the contro##ed proce happen to 3e re#ati,e#y #*ih1 the P.' a#orithm can 3e con-*red to ta$e immediate and dramatic action "hene,er a random dit*r3ance chane the proce ,aria3#e or an operator chane the etpoint Then the P.' a#orithm m*t app#y more coner,ati,e correcti,e e
P.' Contro##er T*nin

P5H&778 T&H

9999ec

(ma0) T'H 7

ec

(min)

For the econd e0periment "hich i contro##er t*nin "a r*n to de,e#op optimiKe the c#oed #oop repone for proce mode# "ith the a3ence of




peci-ed dit*r3ance or et+point chane The F.C9& in :an*a# (:) mode "ith the et point SH26$;hr The P.' ,a#*e "a ad4*ted "ith P5H a3o*t &7781 T.H9999 ec (ma0) and T'H7ec (:in) Then1 the : in F.C9& "a ad4*ted *nti# the P read c#oe to the et point 26$;hr The F.C9& "a tranferred to A*to (A) mode The contro##er need to 3e *re in the A*to mode .n thi tep1 the et point p*#e dit*r3ance "a introd*ced The C#oe Loop :ethod "a repeated "ith a decreain P5 *nti# *niform (contant amp#it*de) oci##ation i o3er,ed .n thi e0periment1 the ,a#*e P5 "a decreae at 67 8 "hich i contant amp#it*de The ditance 3et"een ad4acent pea$ "a a*med 2 mm After the contant amp#it*de appear1 the Tn need to 3e ca#c*#ated 3y *in thi form*#a Tn H mea*red period in mm 0 77 H Chart peed e 677mm;hr

& 0 77 H 2 677

T. H Tn H 2 H  &2

&2

So1 the P5J H 2 0 P5J H 2 0 67 H &778

5aed on the ca#c*#ation1 the ne" P.' ,a#*e i P5H &7781 T.H ec and T'H7ec (:in) A ma4or contro# ytem that ma$e the e



An on;o< contro##er on#y ha t"o o*tp*t1 either f*## ma0im*m or f*## minim*m For thi ytem it ha 3een determined that "hen the mea*rement fa## 3e#o" the et point1 the ,a#,e m*t 3e c#oed to ca*e it to increae Th*1 "hene,er the ina# to the a*tomatic contro##er i 3e#o" the et point1 the contro##er o*tp*t "i## 3e &778 A the mea*rement croe the et point the contro##er o*tp*t oe to 78 Thi contin*o* oci##ation may or may not1 3e accepta3#e dependin *pon the amp#it*de and #enth of the cyc#e Rapid cyc#in ca*e fre*ent *pet to the p#ant *pp#y ytem and e0cei,e ,a#,e "ear On #are capacity proce1 *ch a temperat*re ,at1 the #are capacity ca*e a ,ery #on time contant Therefore1 the mea*rement can chane on#y ,ery #o"#y The diad,antae *in /ie#er Nicho# t*nin method are i,e ,ery arei,e t*nin1 ome tria# and error and a#o proce *pet The diad,antae *in /ie#er Nicho# t*nin method are i,e ,ery arei,e1 ome tria# and error and a#o t*nin proce *pet

Fi*re 6? P.' Contro##er T*nin 5.0

CONCLUSION

I


The e0periment conit dein a P.' contro##er *in /ie#er Nicho# c#oed #oop method1 identifyin the important component of the air !o" contro# ytem and to carry o*t the tart+*p proced*re ytematica##y in a !o" contro# .n addition1 there are t"o indicator that *ed in thi e0periment "hich are Contro# of Ga F#o" Sytem and P.' Contro##er T*nin 5y decreain the ,a#*e of P51 the oci##ation "i## increae and the repone dampin e0treme#y for the econd tria# at &678 compared to 2778 of et point A ma4or contro# ytem that ma$e the e
6.0 REFERENCES

eF*nda1 .nc1 27&&Ori-ce F#o"meter Ca#c*#ator Mon#ine A,ai#a3#e at? http?;;"""ef*ndacom;form*#ae;!*id;ca#cori-ce!o"metercfm MAcceed &2 Octo3er 27&& >a*en F1 27&7 /ie#er+Nicho#B C#oed+Loop :ethod Mon#ine A,ai#a3#e at? http?;;techteachno;p*3#ication;artic#e;Knc#oed#oopmethod;Knc#oed#o opmethodpdf MAcceed &2 Octo3er 27&&

9


7.0 APPENDIXES

GAS FLOW (AF 922) PLANT CONTROL DN.T

&7

process control experiment

Recommend Documents