- G.W. Lockwood (Schlumberger Well Services) | D.E. Cannon (Schlumberger Well Services)
- Document ID
- Society of Petroleum Engineers
- Journal of Petroleum Technology
- Publication Date
- February 1982
- Document Type
- Journal Paper
- 413 - 418
- 1982. Society of Petroleum Engineers
- 5.5.11 Formation Testing (e.g., Wireline, LWD), 4.3.4 Scale, 2.4.5 Gravel pack design & evaluation, 1.6 Drilling Operations, 1.6.10 Running and Setting Casing, 5.6.9 Production Forecasting, 1.2.3 Rock properties
- 1 in the last 30 days
- 593 since 2007
- Show more detail
- View rights & permissions
|SPE Member Price:||USD 5.00|
|SPE Non-Member Price:||USD 35.00|
The integration of old interpretation techniques with recent developments in openhole logging makes it possible to approximate closely a well's production before completion. Watercut predictions can be derived from crossplotting accurate values of water saturation and effective porosity. Flow rates can be obtained from the pressure data of wireline testers. Techniques and examples are discussed.
There always has been a need to know how much oil and/or water a well will make prior to setting casing. As prices increase, it also becomes very important to know whether a well is producing up to its potential. If it is definitely below its potential, then remedial action may be called for. Through a combination of interpretation techniques a good approximation of water cut (WC) and flow rate is possible before a well is completed.
WC is obtained by a Morris and Biggs technique using effective porosity (phi e) and water saturation (Sw) as input values. Phi e and Sw values uncorrected for lithology, shale, and hydrocarbon effects can be used, but any errors in WC prediction will be amplified. The best inputs come from computer-processed data.
Flow rates can be calculated from pressure and permeability data used in conjunction with fluid characteristics data. These data can be obtained from drillstem or production tests, however, both can be lengthy and expensive if there are multiple zones. A more rapid and often more economical source of this information is a wireline test. A wireline tester can make an unlimited number of pressure tests per trip in the hole. It records pressure vs. time. These data are processed through drawdown and buildup techniques to give permeability (k). To convert k into flow rate requires knowledge of the fluid's characteristics, which can be obtained from several methods.
Once predicted flow rates and water cuts are known, completions can be evaluated for optimal recovery.
In 1967, Morris and Biggs presented a technique for calculating WC. It was shown that a crossplot of Sw vs. phi e will describe a hyberbolic curve when grain size is fairly constant and Sw is at irreducible water saturation, Swirr (see Figs. 1 and 2). From this relation, =C. If grain sizes vary, C will vary. By definition, points on the C curve will not make any water. Points plotting above the C curve indicate that water will be produced. The farther from the C curve, the higher the expected WC.
Establishing the C curve can be difficult if there are no zones at Swirr. In those cases there are two ways to establish C. One is from other nearby wells, if the zone is consistent. The second is use of charts, such as Fig. 4. In this method, one enters the figure with phi e and permeability k and reads Swirr (or C). Phi e is available from computed logs. Obtaining a value k is discussed under Flow Rates. If the values obtained from the figure are close to your plot of Sw vs. phi e, it implies you have at least some points at Swirr.
Quantifying WC is done by graphical application of the Morns and Biggs chart solutions to WC (Fig. 5). WC predictions for gas zones are not discussed because of insufficient data.
The easiest way to handle the WC computations is by computer. Software is being tested but is not yet commercially available. For the examples described in this paper, a manual technique was used. A WC scale for the appropriate oil gravity was constructed on the basic phi e vs. Sw plot. First, the lowest C curve to fit the data (the 0% WC curve) was established.
|File Size||353 KB||Number of Pages||6|