ABANDONMENT TEST AT CARRESSE - Brouard Consulting

ABANDONMENT TESTAT CARRESSETechnical ClassSMRI Fall Meeting – Leipzig, GermanyOctober 3, 20101MAIN OUTLINE• Main issues in salt-cavern abandonment• Carresse SPR2 cavern• In-situ measurements• Numerical computations• Computations performed by Ecole Polytechnique-BC• Computations performed by IUB• Main Conclusions21

MAIN ISSUESIN SALT-CAVERNABANDONMENT3PRESSURE BUILD-UP INA CLOSED CAVERNfracture(After Van Sambeek, 1990) Vauvert (Bérest et al., 1979)42

MAIN CONCERNcavern sealed1. Why cavern pressure increases?2. How can we predict cavern long-term behavior?54 MAIN PHENOMENALEAKS• Rock salt behaves as a fluid• Brine warming is a very slow process• Salt permeability is exceedingly small• Leaks can be non negligible63

Cavern compressibilityCavern pressure increase/decrease7FACTORS CONTRIBUTING TO PRESSURE BUILD-UP• SALT CREEP → CAVERN CLOSURE• BRINE WARMING → THERMAL EXPANSIONFACTORS CONTRIBUTING TO PRESSURE RELEASE• ROCK SALT PERMEABILITY → BRINE MICRO-PERMEATION• LEAKS (through casing, casing shoe, at well-head)84

ONE CAN DISTINGUISH 2 MAIN CASES1. Brine thermal expansion can be disregarded- Cavern is not ttall → final pressure smaller than geostaticti- Cavern is tall → overpressure possible at the roof2. Brine thermal expansion cannot be disregarded- Cavern is deep → it is necessary to wait for several years- Cavern is shallow → even a very small permeabilitycan prevent too large a pressure build-up9KEY POINT FOR A LONG-TERM ABANDONMENT TESTAccurate pressures and temperature measurementsare not enoughThe in situ test must be able to prove that there wereNO LEAKSduring the test, or tiny leaks that can be precisely measuredor back-calculated.105

CARRESSE SPR2– A small and shallow cavern –Not at thermal equilibrium11Carresse126

4 shallow caverns at CarresseSPR1SPR2OverviewSPR4SPR3VerticalCross-sectionSPR1SPR3SPR2SPR413SPR2 CAVERN00 ft)313 m (1009000 m 3(56,600 bbls)147

• SPR1, SPR2 & SPR4 were formerly used to store LPGSPR2geometricaltraps• Trapped propane was recover from SPR2 in 2003[de Laguerie et al., SMRI Fall Meeting, Berlin, 2004]22 metric tons of propane recovered15SPR2 CAVERN IS NOT AT THERMAL EQUILIBRIUMTemperature log20023 months0.58°C/year20021.1 °CT18.4 °CT 19.5 °CtMeasurement performed in 2002168

ONE CAN DISTINGUISH 2 MAIN CASES1. Brine thermal expansion can be disregarded- Cavern is not ttall → final pressure smaller than geostaticti- Cavern is tall → overpressure possible at the roof2. Brine thermal expansion cannot be disregarded- Cavern is deep → it is necessary to wait for several years- Cavern is shallow → even a very small permeabilitycan prevent too large a pressure build-up17IN SITUMEASUREMENTS189

Cavern pressure evolutionWELL KNOWNCavern compressibilityBrine thermal expansionNOT WELL KNOWNSalt creepSalt micro-permeationLeaksLONG-TERM FEM COMPUTATIONS19Cavern pressure evolutionWELL KNOWNCavern compressibilityBrine thermal expansionNOT WELL KNOWNSalt creepSalt micro-permeationLeaksLONG-TERM FEM COMPUTATIONS2010

2 years0.5 MPaSyracuseNancyBrussels21+1 MPaJune 20042211

WELL STATUS FROM APRIL 22 TO SEPTEMBER 15 20057 mof oil23LEAK-DETECTION SYSTEM2412

WELLHEAD PRESSURESSeptember 15, 2005 oil injectionJune 2004June 2006April 200525WELLHEAD PRESSURES FROM SEPTEMBER 15 2005 TO MAY 20062 months2613

WELLHEAD PRESSURE FILTERING1 month~2 hours lag27DETERMINATION OF CAVERN PRESSURE EVOLUTIONPressures at well headCavern pressure• columns composition• temperature log• oil/brine compressibilities• leaks2814

CAVERN PRESSURE EVOLUTION AS COMPUTED1 st stepNO OIL4 th step2 nd step3 rd step29Cavern pressure evolutionWELL KNOWNCavern compressibilityBrine thermal expansionNOT WELL KNOWNSalt creepSalt micro-permeationLeaksLONG-TERM FEM COMPUTATIONS3015

CompressibilitytestV33.97 m /MPa31SALT ELASTIC PARAMETERS BACK-CALCULATEDBC-Ecole PolytechniqueIt was assumed: 0.25IUBIt was assumed: 0.35FEM fittingFEM fittingE 16,500 MPaE 17,800MPa3216

Cavern pressure evolutionWELL KNOWNCavern compressibilityBrine thermal expansionNOT WELL KNOWNSalt creepSalt micro-permeationLeaksLONG-TERM FEM COMPUTATIONS33DETERMINATION OF CAVERN TEMPERATURE EVOLUTION1 st task: Determination of natural rock temperature T 2 nd task: Fitting of cavern temperature measurementAccurate prediction of brine thermal expansionduring and after the test3417

100 mSPR2SPR3700 m300 m35DETERMINATION OF SPR2 NATURAL ROCK TEMPERATURESPR2310mT 19.5 °CSPR3700 m3618

SPR2 TEMPERATURE MEASUREMENT IN 2002reference18.35°C – 0.58°C/year37Cavern TemperaturemeasurementT 19.5 °CAccurate predictionactualvirtualTimeJuly 20023819

CAVERN TEMPERATURE FITTING (FEM computation, LOCAS)39CAVERN PRESSURE EVOLUTION WITH/WITHOUT BRINE HEATINGHeating+creep>permeation+leakspermeation+leaks>creepBrine heating is the main phenomenon leading to cavern pressure increase4020

ETREZ EZ53 TEST (1997-1998) – NO BRINE HEATINGpermeation>creepcreep>permeation41Cavern pressure evolutionWELL KNOWNCavern compressibilityBrine thermal expansionNOT WELL KNOWNSalt creepSalt micro-permeationLeaksLONG-TERM FEM COMPUTATIONS4221

LEAKS DETERMINATION43oil injection(Sept. 2005)Pressures at well headPressure difference3 rd step4 th stepoil injection(April 2004)4422

1 year70 hPa ~ 1 psi ≡ 1.8 meter ~ 5.9 ft3 rd step 4 th stepMaximum leak:V70 hPa in 1 year 0.397 liters/hPaAv. leak < 0.08 liters/day ~ 0.02 gal/day45WH pressure differenceWH temperaturewintersummer4623

Cavern pressure evolutionWELL KNOWNNOT WELL KNOWNCavern compressibilityBrine thermal expansionSalt creepSalt micro-permeationLeaksLONG-TERM FEM COMPUTATIONS47SALT PARAMETERS TO BE DETERMINED: Salt creep parametersStationnary ti creepTransient creepNorton-Hoff parametersA, nQ , RMunson-Dawson parametersA , n , m, K , , c1 1o+ reverse creep Slhd Salt hydraulical lparametersKTHESE PARAMETERS MUST BE BACK-CALCULATEDAND/ORA SENSITIVITY STUDY MUST BE PERFORMED4824

Cavern pressure evolutionWELL KNOWNCavern compressibilityBrine thermal expansionNOT WELL KNOWNSalt creepSalt micro-permeationLeaksLONG-TERM FEM COMPUTATIONS49Finite Element ComputationsbyLMS&Brouard ConsultingBack-calculation of salt parameters5025

EXAMPLE OF SPR2 MESH4109 nodes – 7794 elements51STATIONNARY PARAMETERS FITTED FOR TWO PERIODSAverage pressure difference< 10 hPa (0.15 psi)5226

n = 5lt permeabilitySalNorton-Hoff lawQ n Aexp RT Norton A parameter5·10 -20 m²4·10 -20 m²FITTING SENSIBILITY HAS BEEN INVESTIGATEDK 410 m20 2No orton n parameterSet #1Set #2Norton A parameter5427

MAIN FITTING RESULTS20 2 Salt intrinsic i i permeability:K 4 510 m Salt stationnary creep:Q n Aexp RT (Norton-Hoff law)-nA (/MPa -year)nQ R(K)Example of2 good sets of parameters:Set #1 2.5 2.5 4100Set #2 7.8 5 410055Norton-Hoff set #1hydK salt410 m20 2n 2.52.5A 2.5 /MPa -yearQ R 4100 K196420062238Set #15628

Norton-Hoff set #2hydK salt410 m20 2n 55A 7.8 /MPa -yearQ R 4100 KSet #257Norton-Hoff set #1hydK salt410 m20 2n 2.52.5A 2.5 /MPa -yearQ R 4100 KSet #15829

Norton-Hoff set #2hydK salt410 m20 2n 55A 7.8 /MPa -yearQ R 4100 KSet #259Finite Element ComputationsbyIUBSensitivity study on salt parameters6030

IUB STUDIES• data analysis• calculation of equilibrium pressure after sealing• determination of the most reliable parameter set• validation against test observations61s, %/d1,0E-01No lab test possibleCreep rate1,0E-021,0E-0310E04 1,0E-042SPR 2-7-1SPR 1-8-1SPR 2-8-1SPR 1-9-2SPR 3-1-11,0E-05LUBBY2 (all tests)1 10 1001 eq , MPaDeviatoric stress6231

CALCULATION MODELdepth, m-297.0-303.0-305.0-307.0-313.0r = 1.36 mr = 6.74 mr = 6.00 mr = 18.10 m-321.363Results − Long Term PressureAssumed ModelLong Term PressureIncrease (MPa)reference 1.3higher creep ability 2.6lower creep ability 0.2higher permeability 0.9lower permeability 2.6hydrostatic far field p 0 0.9WIPP site far field p 0 1.76432

Norton-Hoff set #1hydK salt410 m20 2n 2.52.5A 2.5 /MPa -yearQ R 4100 K20061964IUB2.6 MPaSet #165PHENOMENA WHICH ARE ACTIVE DURING THE TESTPressureincreasePressuredecreaseBrine thermal expansion ≈1 liter/dayCavern closure (creep) ≈ 0.4 liter/dayBrine micro-permeation ≈ 0.6 liter/dayLeaks ≈ 0.1 liter/dayPHENOMENA WHICH ARE ACTIVE ON THE LONG TERMPressureincrease(MIT: ≈ 400 liters/day)Cavern closure (creep) ≈ 0.5 liter/dayPressuredecreaseBrine micro-permeation ≈ 0.5 liter/day6633

CONCLUSIONS A 2-year long abandonment test on a small and shallowcavern has been performed at Carresse, France. Parameters back-calculation, a sensitivity study, andlong-term simulations have been performed. The existence of an equilibrium pressure far belowgeostatic pressure has been confirmed. It has been proven that this cavern can be safely sealed. Final report is available for members on SMRI Website.6734