Sand Production Prediction
- W.R. Moore (BJ Services Co.)
- Document ID
- Society of Petroleum Engineers
- Journal of Petroleum Technology
- Publication Date
- November 1994
- Document Type
- Journal Paper
- 955 - 955
- 1994. Society of Petroleum Engineers
- 3.2.5 Produced Sand / Solids Management and Control, 1.6.9 Coring, Fishing, 5.3.2 Multiphase Flow, 2.2.2 Perforating, 5.6.1 Open hole/cased hole log analysis, 5.6.9 Production Forecasting, 2.4.3 Sand/Solids Control, 5.2 Reservoir Fluid Dynamics
- 4 in the last 30 days
- 1,658 since 2007
- Show more detail
- View rights & permissions
|SPE Member Price:||USD 1.00|
|SPE Non-Member Price:||USD 2.00|
Many investigators over the years have researched the causes of sandproduction and searched for a reliable means to predict it. Sand productionprediction is important because of the safety, environmental, and operationalconcerns involved when produced sand particles fill and plug the wellbore,erode downhole and surface equipment, and increase operating expense.Currently, no method of sand production prediction is universally regarded asaccurate and reliable within the industry.
A number of prediction models have been developed to identify completionsthat may be expected to produce sand. Earlier attempts to develop predictiontechniques included statistical models, numerical models, mechanical propertieslogs, sand strength logs, and core studies. Often the individual attempts todevelop a predictive model were specific to the type and locale of thereservoir being studied; i.e., water production, pore pressure depletion,perforation geometry, pressure drawdown caused by skin effects, and a varietyof other critical parameters were not always considered.
The dilemma of an independent U.S. operator faced with a completion decisionregarding sand control is not that different from the problems considered bylarge North Sea production companies. Quite often, the most cost-effectivemethod to determine the need for sand control is by analogy from data collectedfrom offset wells. Complex 3D numerical modeling in concert with extensivelaboratory analysis of core and log data is not always economically practicalbut is the most technically correct method with an acceptable degree ofaccuracy when properly performed.
Historically, much attention has been given to sand production prediction.The vast differences and complexities observed between the models andtechniques that have been developed suggests a multitude of engineering andgeologic parameters to be considered (Table 1).
Morita and Boyd cataloged and analyzed five typical sand problems observedin the field that were induced by (1) unconsolidated formations, (2) waterbreakthrough in weakly to moderately consolidated formations, (3) reservoirpressure depletion in relatively strong formations, (4) abnormally high lateraltectonic forces in relatively strong formations, and (5) sudden changes in flowrate (cyclic loading) or high flow rate. They developed a core-based completionguide on the premise that completion performance in a weak formation issignificantly affected by near-wellbore rock strength and permeability.
Morita and Boyd determined that the strength of a reservoir rock variessignificantly depending on stress levels and that sandstones deform nonlinearlywith the nonlinear character varying with the stress rate. In other words,sandstone failure is a dynamic process that is dependent on changingnear-wellbore stresses that are affected by some combination of the 16 factorsin Table 1. An evaluation of the completion at one point in time withoutextrapolation into the future will not accurately predict the conditions underwhich failure will occur.
Sonic, density, and neutron logs relate the well under study to an activesand-producing well. Openhole logs must be calibrated to a known sand producerfor an analogous comparison to the well being studied. The problem is that thewell must be taken to the point of formation failure before the comparison canbe made, and no technique exists for forecasting reservoir performance into thefuture. Additionally, the effects of pressure depletion, water production(multiphase flow), and additional pressure drops caused by skin effects are notconsidered. Wells located offshore in the Gulf of Mexico are a good example ofthis decision criteria.
The sand production prediction technique that will be most effective for agiven area is one in which field observations, laboratory experiments, andtheoretical modeling are all integrated for the formation under study.
1. Veecken, C.A.M. et al.: "Sand Production Prediction Review:Developing an Integrated Approach," paper SPE 22792 presented at the 1991SPE Annual Technical Conference and Exhibition, Dallas, Oct. 6-9.
2. Morita, N. et al.: "Parametric Study of Sand Production Prediction:Analytical Approach," paper SPE 16990 presented at the 1987 SPE AnnualTechnical Conference and Exhibition, Dallas, Sept. 27-30.
3. Morita, N. and Boyd, P.A.: "Typical Sand Production Problems-CaseStudies and Strategies for Sand Control," paper SPE 22739 presented at the1991 SPE Annual Technical Conference and Exhibition, Dallas, Oct. 6-9.
4. Ghalambor, A. et al.: "Predicting Sand Production in U.S. Gulf CoastWells Producing Free Water," JPT (Dec. 1989) 1336.
|File Size||85 KB||Number of Pages||1|