An LES analysis of industrial flare performance under high wind conditions

/Qk C-SAFE U UNIVERSITY OF U T A H A n L E S A n a ly s is o f In d u str ia l _____ _______________ -i_____ __________________________ ______ ._______________ ____ _____ ,__ ._____ _________ F \n re P z r r o r r n n n z z u n d z r High WindConditions 7 -8 N o v 2 0 0 5 AFRC, A tla n ta P h ilip J. S m ith P ro fe s s o r & C h a ir D e p t, o f C h e m ic a l E n g in e e rin g T h e U n iv e rs ity o f U ta h Seshadri Kumar, Jennifer Spinti, P ad m ab h u sh an a Desam, Stanislav Borodai The field h as reached a threshold at which better organization b eco m es crucial. New m eth o d s of verifying and validating complex c o d e s are m andatory if computational scien ce is to fulfill its prom ise for scien ce and society. THE U IN IV E R S IT Y OF U T A H n> nw» — ******* . T f 7 -8 N o v 2 0 0 5 ^ * - i ^ **»——' - . •»»•> r r t r r i T r .r r ^ . T 4' ~ <*towu >!»* ..nun, ~U«. »— *»»■ .TT T*1 . '‘. T..?* „ a m !■■■(•( al la ataaaMat —a* ■* np-i- t ■ I f ii i Ui” ^?. • r U *** il | ^^ a a a a ^ ------- , ----—------------- - J|..W .«. M^aiH I i <i mT."'*4 ■•UaUa ■ a rf ■I■ u MM «. w»a«H ■ i a .M .’tUm m , t* h !** — r— -------1— —~ *------ --------- “ P h ilip J ^ P r o fe s s o r & C h a ir D e p t, o f C h e m ic a T h e U n iv e rs ity o f Seshadri K Padm abhu __ L^ |n> 7| . .. - r X T *"' *» ■ ! i 1-- ‘ »«- Kp. M »—• t**nnal m m* *«■■■■ tmr | AFRC, A tla n ta “T h e p u r p o s e of c o m p u t i n g is in si ght , n o t n u m b e r s ” (1961) SQk “T h e p u r p o s e of c o m p u t i n g n u m b e r s is n o t y e t in s i g h t ” ** ^ A F E u UNIVERSITY OF UTAH (1997) I Richard W. Hamming (1915-1997) r IcJTbl r z i f i O f l f J s J f High WindCond P h ilip J. S m ith P ro fe s s o r & C h a ir D e p t, o f C h e m ic a l E n g in e e rin g T h e U n iv e rs ity o f U ta h Seshadri Kumar, Jennifer SJ. P ad m ab h u sh an a Desam, Sta> after Oberkampf & Trucano, Sandia National Labs C o m p lete S y stem h e a t f tu x & e m i s s i o n s f r o m in d u strial fla r e s S u b sy ? tem C a s e s plum e v e lo c ity h e a t flu x & em m issions p lu m e s h a p e B enchi lark C ases b u o y a n c y d riv en fla m es ( lik e T N F d a t a ) S an d ia TNF F la m e s U n it P r o b le m s CFD a lg o rith m su b g rid tu rb u len ce re a c tio n m odel m ix i n g m odel rad iatio n m odel | la m i n a r flo w w ith | c om plex g eo m etry C o m p lete S y stem Flare S h a p e Data f la r e w in d o b s e r v e v s. f la r e i m a g e s d y n a m i c s m o m . (v a lid a te w tu n n e l ra tio s h a p e o f f la r e s c r o s s w in d ) unit Problems | la m i n a r flo w w ith | c om plex g eo m etry CFD a lg o rith m K in e tic M e c h a n is m s I 1 L E S g a s p h a s e chem ical kinetic m o d el 1 H 1 Bj so o t m odel ( w i th o p t i c a l p r o p e r t i e s ) I I s im u la tio n t J o h n s o n & K ostiuk, C o m b . & F lam e, 123: 189-200(2000) K\ ■ * short exposure photograph long exposure photograph \ C a se A: J e t vel = 2.07 m/s Wind vel = 0.7 m/s R = 5.29 C a s e O : J e t Vel = 2.09 m/s Wind vel = 4.09 m/s R = 0.15 Complete System h c a tfk i J plum e velocity — A lberta S o u r G a s Flares am pi. fu ll s c a l e lid a r c o m p o s itio n d a t a v id e o o f f la r e K in e tic M e c h a n is m s so o t m odel ( w i th o p t i c a l p r o p e r t i e s ) A lb e r ta R e s e a r c h s o u r g a s flare C o u n c il (A llan C h a m b e r s ) A lb e r ta (A llan C h a m b e r s ) s o u r g a s flare (A llan C h a m b e r s ) D o m a in S ize R u n T im e: S im u la tio n T im e: R e a c tio n M odel M ix in g M o d e l C h e m . M ech C o u n c il s o u r g a s flare M e sh R e so lu tio n C o m p u ter: R e s e a r c h L L N L - f r o s t (IB M S P ) 48 hours on 320 p ro cesso rs 5 .9 s o f s im u la tio n tim e s te a d y lam in ar fla m e le ts PDF (c lip p e d g a u s s la n ) REI m e c h a n i s m w i t h s u l f u r (69 s p e c ie s , 7 7 4 re a c tio n s} A l b e r t a R e s e a r c h s o u r g a s flare D o m a in S ize: M e sh R e so lu tio n : C o m p u ter: R u n T im e: S im u la tio n rim e : R e a c t i o n M o d el: M ixing M o d e l: C h e m . M ech: 300x210x96 L L N L - f r o s t (IB M S P ) 48 ho u rs on 320 p ro ce sso rs 5 .9 s o f s im u la tio n tim e s te a d y lam in ar fla m e le ts PD F (clip p ed g a u s s la n ) REI m e c h a n i s m w i t h s u l f u r (6 9 s p e c ie s , 7 7 4 re a c tio n s} R e s e a r c h s o u r g a s flare D o m a in S ize: C o m p u ter: R u n T im e: S im u la tio n rim e : R e a c t i o n M o d el: M ixing M o d e l: C h e m . M ech: (A llan C h a m b e r s ) 2 5 .0 m x 1 7 .5 m x 8 .0 m A lb e r ta M e sh R e so lu tio n : C o u n c i l 2 5 .0 m x 1 7 .5 m x 8 .0 m 300x210x96 L L N L - f r o s t (IB M S P ) 48 ho u rs on 320 p ro ce sso rs 5 .9 s o f s im u la tio n tim e s te a d y lam in ar fla m e le ts PD F (clip p ed g a u s s la n ) REI m e c h a n i s m w i t h s u l f u r (69 s p e c ie s , 774 re a c tio n s) C o u n c il (A llan C h a m b e r s ) e f f e c t o f h i g h w i n d s p e e d i n c r e a s e d c o m b u s tio n stab ility (“w a k e s t a b i l i z e d ” ) e f f e c t o f h ig h w in d s p e e d i n c r e a s e d c o m b u s tio n stab ility (“w a k e s t a b i l i z e d ” ) 1 .0 0 Combustion Efficiency : 4 B P R W 1 r 0 .9 5 0 .9 0 0 .3 5 Ret RftittinnEjjl*ifmy•>(IniuMritlthirds T h e P i r y v t f ii r » ! th e P tt \f. I (i V v h i l , II C’ Davit, P I I W Koftiufc. J II. IN*hl. K I Schwartz. S k V«clKr)!. VI 0 .8 0 Mroshor aikl P M W alsh < orobuxtion ( .iiuib •• V V oncom cf, B e . V p ic ro lv r 2 0 0 J 0 .7 5 101 102 1fr’ 10° 10’ 103 < p V V ( P v x ,™ < 10! 10* 10s 10f d ec rea sed co m b u stio n efficien cy L E S & f la r e p e r f o r m a n c e - LES re s o lv e s large s c a l e d y n a m ic s validation in fra stru c tu re => se e k in g q uantitative predictability - co m b u stio n inefficiency u n d e r high wind a f l a m e extinction by c ro s s w in d in d u c e d strain - f la r e p e rf o rm a n c e scaling =>s im u la tio n