Ig Nobel Prized Research

Transcription

1 Ig Nobel Prized Research 2010 Chemistry Prize Eric Adams (MIT) Scott Socolofsky (TAMU) Steve Masutani (U. Hawaii) British Petroleum for disproving the old adage that oil and water don t mix Selected other winners: Engineering: Remote controlled whale snot collector Health: Roller coaster asthma therapy Peace: Swearing relieves pain Management: Organizations should promote people randomly

2 Industry Technical Advisory Committee The Fate of Deep Sea Plumes E. Eric Adams Massachusetts Institute of Technology Cambridge, MA USA October 24, 2016

3 Collaborators Scott Socolofsky, Texas A & M Michel Boufadel, NJIT Steve Masutani, U. Hawaii Oistein Johansen, SINTEF Per Johan Brandvik, SINTEF Elizabeth North, U. Maryland

4 DWH spill site & data collection In-situ measurements of CDOM, temp, conductivity, etc collected by cruises w/o intrusion w intrusion o Sample period: ~3 mo. o Sample area: 7km o Total # of casts: 266 Terra Satellite, NASA, as of May 24, 2010 o Casts w/ intrusion: 111 (87)

5 Intrusion Types

6 # of casts Variation with Time and Space Type A Type B Type C Type D Type E 0 05/12 05/22 06/01 06/11 06/21 07/01 07/11 Date A B C D E

7 gas released in linear stratification Density stratification is caused when lighter (warmer) water overlies heavier (colder) water. Light gas and heavy seawater rise to a level of neutral buoyancy causing oil and seawater to separate from the gas and intrude laterally. Gas bubbles continue to rise, causing plume to restart, with lesser quantities of oil Socolofsky & Adams (2003)

8 oil and gas released in a current Gas bubbles create plume, but currents blow oil and entrained seawater downstream, leaving gas to rise separately. Socolofsky & Adams (2002)

9 Plumes in quiescent ambient Plume behavior depends on B, N, Us, z B = Qg r - r w r w N = g r / r z U c = (BN ) 1/4 L c Q c ( B N ( B 3 3 N ) 1/ 4 5 ) 1/ 4 U s = droplet slip velocity Socolofsky & Adams (2005); Crounse et al. (2007); Seol et al., (2009)

10 Stratification to Crossflow Domination Increasing U a Stratification dominants Cross-flow dominates h To /L c = 2.9exp[-(U s /U c -1) 2 /27] h T = h To exp[-ku a /U c ] h s = 5.1B/(U a U s 2.4 ) 0.88 U a /U c = (U c /U s ) 2

11 Observed and Predicted Trap Heights h T /(B 1/4 /N 3/4 )= 2.9exp[-(U s /(BN) 1/4-1) 2 /27]*exp[-kU a /(BN) 1/4 ] (U a = 0) Socolofsky, Adams, Sherwood (2011)

12 Variation in current speed, U a A Origin B h S Origin h i h T h i h P Displaced origin x h total U a U a x o 4 z 1.2 x c Wang and Adams (2016) 3 1 h T /(B/N 3 ) 1/4 z 2 1 f 2 = h T /(f 1 l c ) z 6 8 U s /(BN) 1/ U a /U c

13 Variation in buoyancy flux, B Small jets at the kink during first 1.5 months Kink jet Slugging Oily flow Gassy flow

14 Variation in local stratification, N Temporal variation of N at source Spatial variation in N as reflected in CDOM casts

15 Variation in droplet size, Us

16 Sensitivity relative to reference case N B U a U s

17 Secondary intrusion height, h T2 h T ~ B1/4 N 3/4 h T 2 h T1 = (B 2 / B 1 )1/4 (N 2 / N 1 ) 3/4 B 1 = B o + B g B 2 = B o r ~ z 2 N ~ z 1/2 h T 2 h T1 = 0.56 h T2 /h T Ave of obs. = h T2 h T Cast #

18 Secondary Intrusions, cont d Less likely with height (Plumes more likely to be crossflow dominant U a /U c > (U c /U s ) 2 U a > B 3/4 N 3/4 /U s 2 Less likely to be observed (lighter fractions have dissolved by first intrusion)

19 Yang et al. (2016) Additional details with CFD

20 Plume Classification (Quiescent Conditions) s Type 1a* Type 1b* Type 2 Type 3 U N < < U N < < U N < < U N U N = u s /(BN) 1/4 ; B = plume buoyancy flux; N = stratification frequency Smallest droplets are broadcast most widely Chan et al. (2014); Socolofsky & Adams (2003); Asaeda & Imberger (1993)

21 Initial Jet Break-up In turbulent regime, droplet size not directly scaled by outlet Tang & Masutani (2003) Transition between laminar and turbulent regimes depends on We = ru o2 D/s

22 Oh Atomization Large drops We = ru o2 D/s DWH Re Tang, Gorgas & Masutani (2003)

23 SINTEF Tower Basin Experiments d 50 /D = 15We/[1+0.8Vi(d 50 /D) 1/3 ] DOR = 0 1:50

24 Numerical (Population) Models Based on a set of differential equations of droplets of various diameters. dn (, t) For each diameter δ, i Birth i Death dt Bandera & Yapa (2011) Zhao et al. (2014)

25 Latent Droplet Behavior unfiltered DOR <~ 1:1000 (elliptical droplets stable for days) filtered DOR >~ 1:250 ( jellyfish disappear in 10s minutes) Nagamine & Masutani (2014)

26 Sigma (m) Application to Deepwater Horizon km >>100 km [>> 100 X non-disp] X m m 750 m 20 km [ X non-disp] m s s 10 1 With dispersant after latent break-up d/ diameter (microns) With dispersant after jet breakup (d/10) No dispersant (d) Chan, Chow & Adams (2014)

27 Vertical Microstructure w e = e 1/2 /N 1/2 Wang, DiMarco & Socolofsky (2016) 27

28 Turbulence Measures Increasing turbulence w e = e 1/2 /N 1/2 Wang, DiMarco & Socolofsky (2016) 28

29 Effect on Droplet Rise w s = w e d e = f -1 (w s ) [Zheng & Yapa (2000); Clift (1978) with Dr/r = 0.85] Disp; w/ LBU Disp; w/ JBU No disp Wang, DiMarco & Socolofsky,

30 Far Field Behavior of Droplets with Degradation Half-life = 1.2 d 3.1 d 6.1 d Infinity North, et al., 2014 North, Adams & Socolofsky (2014)

31 Questions?