dynamic underbalance perforations bring higher productivities than conventional perforations a large...

Download DYNAMIC UNDERBALANCE PERFORATIONS BRING HIGHER PRODUCTIVITIES THAN CONVENTIONAL PERFORATIONS A large scale comparative review from the Tunu Gas Field (Indonesia)

Post on 28-Dec-2015

214 views

Category:

Documents

1 download

Embed Size (px)

TRANSCRIPT

  • DYNAMIC UNDERBALANCE PERFORATIONS BRING HIGHER PRODUCTIVITIES THAN CONVENTIONAL PERFORATIONSA large scale comparative review from the Tunu Gas Field (Indonesia)Sebastien Perrier, Tomi Sugiarto, TOTAL E&P INDONESIEEuropean and West African Perforating Symposium 2012

    Presentation title - Place and Country - Date Month Day Year

    EWAPS 2012

    Tubingless completionsPerforations are addedby light intervention barges (electric lines)Multilayer thin sandstone Reservoirs

    Gas + condensate

    Deltaic channelsand mouth bars

    10-20% Porosity1-500md Permeability

    What is the TUNU field (Indonesia, PSC Mahakam)?

    Annually: more than 2500 ms of perforations, 600 intervals Large activities and workload continue to increase as field become mature Sufficient data for developing comprehensive perforation analysis

    Presentation title - Place and Country - Date Month Day Year

    EWAPS 2012

    WHAT IS DYNAMIC UNDERBALANCE (DUB)?Dynamic Underbalance (DUB) TechnologyGenerate a large dynamic under balance from modest static under balanced or over balanced pressure.DUB system needs set of charges and additional void space inside the gun (blank section, implosive chamber).DUB system required liquid (water / oil / mud) around the gun when shot.Fast gauge is run in tandem with the gun to measure actual dynamic underbalance (optional, record up to 100,000 samples/second)Unload using coiled tubing and nitrogenFiredResult: pressure drop within perforation intervals by 1500 to 2500 psi during a few milliseconds Objective: optimized clean up of the perforation cavity

    EWAPS 2012

    DOES IT WORK?

    Legend in this presentation:Conventional perforations, without DUBPerforations with Dynamic UnderBalance

    Data from TUNU gas field (2005-2011), operated by TOTAL EP INDONESIE

    Presentation title - Place and Country - Date Month Day Year

    EWAPS 2012

    ARE WE SURE THERE IS NO BIAS?

    There are always some classical biasChoice/selection of best reservoirs for application of DUB? Quality bias?~30 DUB perforations only: luck effect ?

    The question becomes: once we remove quality bias and adjust for the sample size, will DUB still make a difference?

    Presentation title - Place and Country - Date Month Day Year

    EWAPS 2012

    FIRST METHOD: A POLL (13 INNOCENT PAIRS OF EYES)!

    Poll with 13 engineers/Geologists familiar with the Field: Random sets of 50 reservoirs , with all the data used by Tunu field practitioners15 random reservoirs perforated with DUB dissimulated in each set

    1 question: for each reservoir, which perfo. gain (in mmscfd) do you expect?

    (MMscfd)

    Participants guess(average)Real resultConv. perfos2,0 / 6 2,1 / 6Dynamic UB2,4 / 62,8 / 6

    Gain20Mark123456

    Presentation title - Place and Country - Date Month Day Year

    EWAPS 2012

    SECOND METHOD: USING EPIDEMIOLOGICAL METHODS

    Epidemiological methods = classically used to test new drugs or health issues Which steps?Validation of data quality is essentialPerforations without a clear outcome are not considered No perforation job with multi reservoirsReservoirs diagnosed with technical problems or water problems are excludedNeed to have reliable pressure data (cf infill wells)

    The statistical influence of each parameter on differences between results using the same perforation technique must be understood before any comparisonDegree of dependence between parametersInfluence of porosity, res. pressure, etcgun size, static underbalance, brand, etc

    (cf. SPE paper 158083 for extensive details)

    Presentation title - Place and Country - Date Month Day Year

    EWAPS 2012

    SECOND METHOD: USING EPIDEMIOLOGICAL METHODS

    Epidemiological methods = classically used to test new drugs or health issues

    The challenge: Build comparable samples, free of size and quality biasMore than 200 eligible conventional perforations29 flowing DUB perforations

    Two possible approachs: Our original approach Our group of 29 DUB perfo is the reference sampleLets downscale the 200 conv. perforations into random draws of 29 representatives, respecting same res. characteristics as DUB samples

    To get an apple to apple comparison, let s try a comparison between Basket filled with 29 apples and baskets with 29 apples !

    Presentation title - Place and Country - Date Month Day Year

    EWAPS 2012

    RANDOM GENERATION OF 150 COMBINATIONS OF 27-30 CONV.PERFOS, WITH SAME RESERVOIR PROPERTIES AS DUB SAMPLESub-samples randomly created, but matching DUB sample distributions of porosity, pressure, mobility, and thickness

    Presentation title - Place and Country - Date Month Day Year

    EWAPS 2012

    RESULT: AN EXPLICIT INSIGHT IN THE PERFORATION RESULTSDistribution of Results (in mmscfd/m) of the 150 randomly created sub-samples of conventional perforationsAt 90% confidence level: Dynamic UB does bring better productivity results than conv. perfos

    Explicit evaluation of benefit = +10 to +25% initial productivity (3.7-4.3 vs 3.4)

    EWAPS 2012

    CONCLUSIONSIn Tunu gas sandstones, Dynamic UnderBalance makes a difference+10 to +25% initial productivity per meter of perforation

    after removing all possible statistical bias (reservoir quality & sample sizes)

    10 additionnal jobs performed post-study reinforce the quantitative resultsThey still give 4.0mmscfd/d, with similar quality

    Applicability?In gas wells, DUB is limited to wells with a liquid column (provider)Unlikely adapted for tighest reservoirs (operators observation)

    Why not having worked with evaluations of skin ?Too little data to be analysed (see paper)Skin data validity and quality are a major variable, with little control, as associated to well testing procedures

    Independent analysis, and more in SPE158083

    Presentation title - Place and Country - Date Month Day Year

    EWAPS 2012

    AcknowledgementsTotal E&P IndonesieINPEX

    BPMigas - MIGAS

    Thank You / QuestionsPaper # Paper Title Presenter NameSlide *

    Presentation title - Place and Country - Date Month Day Year

    EWAPS 2012

    DYNAMIC UNDERBALANCE SUCCESS Technique Design gun to get optimum dynamic underbalance Design 100psi static UB and estimate the reservoir pressure is enough to lift up the fluid column Clean up the well right after perforation, CT ready to start up in case well unable to flow RIH fast gauge tandem with gun to measure the dynamic underbalance

    Gun Type 2.5 PJO 2506 estimated DUB at bottom gauge: 1560 psi real DUB at bottom gauge was 1620 psiCompletion Data Casing Size: 3.5 (9.2ppf) Min restriction: 2.81 Brine in borehole (1.03 sg) BHP=5120 psiReservoir Data Porosity: 12.95% Pore Pressure: 5211 psi Mobility: 31.6 Interval 3620.5-3623.5m (3m reservoir P131)Result Real DUB at bottom gauge: 1620psi (fast gauge data) Well flow 15.9 MMscfd without unloading job

    **In this presentation, BLACK is for without DUB, Red with DUB

    At first glance, Yes it works:Higher proportion of perforations bringing gains superior to 15mmscfd, with impressive number above 20mmscfd. (left plot)Even corrected from thickness, in gain per meter, higher proportion of good perfos (right plot)*Now the question is: are we comparing APPLE to APPLE, same quiality of reservoirs on both cases

    The simple plot below shows that no.Porosity, that is a major petrophysical indicator of reservoir quality is better for the DUB reservoirs

    There is another possible statistical bias: it is not the same thing (in term of standard deviation) to test 30 times or to test hundreds of times: luck effect is still possible.

    In this presentation, we will remove this two risks of bias, to get conclusions that are not affected by difference in quality or in sample size.*We will propose 2 independent methods:

    This first method is very original, YET very efficient, and also interesting:

    What about a blind test?We will compare gain expectations from engineers familiar with the field, with real gains. Some DUB perforations are hidden in the middle of the sample.13 different engineers, each given 50 randomly selected reservoirs.

    They have to put a gain, then converted into a mark.

    ANALYSIS: The guess on average is 2/6. The real value for those reservoirs was 2.1: WE CONFIRM the INTERVIEWEE HAVE A VALID JUDGMENT, they are experts of their field

    2. The mark the interviewee have given to the hidden DUB reservoirs is 2.4 in blind test, they have identified that these reservoirs had a better quality (confirmation that they are experts !!!)

    3. The KEY thing, is that the real value for DUB is significantly higher that what the interviewee were estimating (2.8 vs 2.4) : this shows that DUB behaves in a better way that what the engineers are used to with conv perfos.

    THIS is a qualitative demonstration that beyond difference in quality, there is an objective difference in results of both techniques.

    However, we propose a very systematic epidemiological study, that will remove also the bias associated to different sizes in samples

    *Slef explaining

    The obejective is to prepare a statistical (epidemiological) comparison.*This part is essential to nderstand well:We have created randomly baskets of 29 reservoirs with all the same properties as our unique 29 sample perfo with DUB.This will be shown on next slide.

    By doing so, we will account for the statistical standard deviation between groups of 29 res.:Although they have same properties, their results will follow a statistical distribution, some groups better than others, etcHow will our DUB sample compare ?(we will see that very few groups of 29 conv res with same properties are as good as DUB)*This slide shows the result of the generation of our sub samples.It is a bitcomplex, but shows how (in grey)