A Method for Pressure Buildup Analysis of Drillstem Tests

Drill-stem tests (DST's) in porous media typically consist of a sequence of production and shut-in periods. Interpretation of DST's is often based on application of the Horner method to the shut-in periods. Such an analysis assumes that the wellbore rate is constant during the production period and that the pressure response during the shut in is affected only by that production period. For many DST's, however, these assumptions are no longer valid and other interpretation methods are required. For example, prior production and shut-in periods and/or pre-test pressures can significantly affect the pressure response of subsequent periods and cause erroneous parameter estimates. In this paper, we present a new DST analytical solution that can account for the effect of pre-test pressures, prior test periods, wellbore temperature variations, and other factors such as wellbore storage and skin on the pressure response. The solution is an extension of the solution presented by Correa and Ramey who derived a DST solution by writing a single time-dependent wellbore boundary condition. Application of the new solution to example DST data shows that estimates of formation pressure from the Horner analysis can deviate from the true formation pressure by several hundreds of psia, whereas permeability estimates can differ by a factor of two or three. The errors in the estimated formation parameters increase as the permeability is reduced, but can be significant even for large permeability reservoirs. These errors are more pronounced for longer pre-test pressure periods and when the pre-test pressures are much greater or smaller than the true formation pressure. The effect of wellbore temperature variations are also shown to be important for DST interpretation in certain cases when the permeability-thickness product is small. Introduction The interpretation of drill-stem tests (DST's) provides important reservoir information for determining the profitability of a particular reservoir or well. The DST can provide good estimates of formation permeability and initial pressure as well as samples of the reservoir fluid. Because DST's are typically composed of a sequence of very short test periods, however, accurate interpretation of DST' s can be challenging. Methods to interpret drill-stem tests have been extensively studied in the literature. Interpretation of DST's is often based on application of the Horner method. Such an analysis assumes that the wellbore rate is constant during the production period and that the pressure response during the shut-in is affected only by that production period. In many cases, however, these assumptions are not satisfied and interpretation of DST's by classical methods may be very inaccurate. For example, prior production and shut-in periods and/or pre-test pressures can significantly affect the pressure response of subsequent periods. The focus of this work is to examine the effect of pre-test pressures, such as those encountered during drilling, logging, or coring, on the interpretation of drill-stem tests (DST). A secondary focus is to examine the effect of borehole temperature variations on the pressure response, which can also impact the DST pressure for certain low-permeability reservoirs. Neuzil and Pickens et al. describe processes and mechanisms associated with well tests in relatively low-permeability formations. The measured formation pressure response in those formations is primarily affected by the following:formation flow parameters and model;wellbore storage, skin and tool compliance;variable pretest pressures;wellbore temperature changes; andcomplex test-sequence design. P. 87^

ABSTRACTA closed chamber test (CCT) is characterized by a variable wellbore storage coefficient, whereas, in a slug test, the wellbore storage coefficient is assumed to be constant. It is shown that for ranges of parameters often encountered in practical cases, a large span of the CCT pressure data can be converted to equivalent slug test pressure data which can then be analyzed by several available slug test analysis techniques. In addition, we derive a new convolution method for analyzing buildup data obtained in a closed chamber drillstem test. We also present a method to analyze slug test data influenced by a step change in the wellbore storage coefficient due to a change in the well-string diameter.

In field data from slug test (decreasing flow rate) DSTS, an initial period of constant flow rate can often be observed. This may be a consequence of critical flow. "Critical flow" means that flow rate is independent of the pressure drop across the flow restriction. In the case of compressible fluid flow, this implies that the flow velocity is equal to the sonic velocity in the fluid. This was suggested by Ramey et al. in 1975. However, critical (constant-rate) flow was not included in their analysis of slug test DST flow period data. A new method which does consider an initial constant flow rate for a slug test was found. Graphs for the type-curve matching method are presented and field data are shown. Introduction Often in drill stem testing, the flow periods are characterized by a pressure trace which increases with increasing time, showing the accumulation of liquid in the drill string. In the expected case, the pressure-time trace is linear, showing constant-rate pressure-time trace is linear, showing constant-rate production. In other cases, the pressure-time trace production. In other cases, the pressure-time trace curves and is concave to the time axis. This shows a decreasing flow rate. In the case that formation pressure is too low to lift a column of reservoir pressure is too low to lift a column of reservoir liquid to the surface, the well may stop flowing before the tester valve is closed. This results because the head of liquid in the drill string becomes equal to initial formation pressure. This sort of test is similar to a pressure transient test called a "slug test" by Ferris and Knowles in 1954. The word "slug" refers to the maximum volume of liquid which may be produced by the time the well becomes static. produced by the time the well becomes static. A similar test involving the cooling of a batch of hot water was reported by Beck, Jaeger, and Newstead in 1956. Cooper et al., in 1967, reported the results of a field test in a static water well from which a float was suddenly removed, giving the appearance of the sudden removal of a quantity of water equal to that displaced by the float. Maier presented an approximate analysis of the equivalent DST presented an approximate analysis of the equivalent DST problem in 1970. problem in 1970.Van Poollen et al., in 1970, and Kohlhass, in 1972, applied the Cooper et al. solution to DST flow period data analysis. Although most current studies period data analysis. Although most current studies reference a study by Jaeger in 1956 which included a sandface resistance similar to the skin effect, most recent works do not include wellbore damage effects. A solution including the skin effect was presented by Ramey and Agarwal in 1972, although use of the solutions was not demonstrated. Papadopulos et al. presented extended results for the zero skin effect presented extended results for the zero skin effect case in 1973. The most complete discussion of DST applications of the slug test was presented in 1975. Three new slug test type-curves were developed for analysis of flow period data, and three field cases were analyzed. This study was included in the Earlougher monograph in 1977, and large-scale type-curves were printed with that monograph. printed with that monograph. Finally, Fenske showed that pressure buildup data for a well produced a short time (with wellbore storage) would follow the slug test type-curves. This is a potent observation. It means that the slug test type-curves may work for the initial shut-in, as well as for flow period data. This establishes the importance of slug test type-curves for DST work whether the test would normally be classified a "slug test" or not. Although not within the scope of this study, slug test type-curves may be used to match the initial shut-in for a DST following a very short initial flow to provide permeability, skin effect, and extrapolated initial pressure. One of the most important observations made from attempts to match field data with slug test type-curves is that there is often an initial constant-rate flow period before the rate begins to decline. Reference period before the rate begins to decline. Reference 1 pointed out this phenomena was related to conventional DST results for liquid flow wherein flow rate was constant for both flow periods: "Before proceeding with the presentation of the 'slug test', it should be pointed out that the conventional DST… contains the seed of a paradox.

This paper was prepared for the 47th Annual Fall Meeting of the Society of Petroleum Engineers of AIME, to be held in San, Antonio, Tex., Oct. 8–11, 1972 Permission to copy is restricted to an abstract of not more than 300 words. Illustrations may not be copied. The abstract should contain conspicuous acknowledgment of where and by whom the paper is presented. Publication elsewhere after publication in the JOURNAL paper is presented. Publication elsewhere after publication in the JOURNAL OF PETROLEUM TECHNOLOGY or the SOCIETY OF PETROLEUM ENGINEERS JOURNAL is usually granted upon requested to the Editor of the appropriate journal, provided agreement to give proper credit is made. provided agreement to give proper credit is made. Discussion of this paper is invited. Three copies of any discussion should be sent to the Society of Petroleum Engineers Office. Such discussions may be presented at the above meeting and, with the paper, may be considered for publication in one of the two SPE magazines. Abstract In a wildcat well or early in the life of a reservoir, drill stem test data is quite often the only means available to estimate reservoir values of transmissibility, effective permeability, well bore damage, well bore permeability, well bore damage, well bore stimulation and productivity. The Horner pressure buildup method is generally accepted as the best means of determining reservoir data from short time tests such as a drill stem test. The Horner method requires that the steady state or straight line portion of the buildup curve be reached. For various reasons, approximately 25% of all drill stem tests run do not meet this requirement. This group of wells are the ones with the highest economic risk involved in making a decision on whether to abandon or complete. A 1971 paper published by Mr. McKinley, "Transmissibility from Afterflow Dominated Pressure Buildup Data", gives a means to Pressure Buildup Data", gives a means to calculate reservoir values when pressure buildup curves are still under the influence of afterflow. A research program was recently completed using drill stern test data to calculate transmissibility and effective permeability by the McKinley method and to compare the results obtained against the Horner method. Fifty drill stem tests were chosen for analysis that had an appreciable afterflow period and steady state period on the same buildup curve. The comparison of results was very close as long as the basic assumptions were not deviated from too far. Most drill stem tests, where it is necessary to use the afterflow buildup method, do deviate considerably from these basic assumptions. Therefore, it is necessary to use a correction factor in interpreting the afterflow method.

Drill Stem Test (DST) describes the dynamic characteristic of the hydrocarbon bearing formation such as wellbore storage, skin effect, permeability, average reservoir pressure and reservoir boundary. The wellbore storage effect and average reservoir pressure help to predict the flowing phase from the reservoir. An effort has been made to analyze the DST conducted in the Kailashtilla field at the depth interval 3261 meter to 3266 meter in well KTL-7. Two sets of pressure profile have been recorded. First conditioning the well for an hour then performed drawdown following pressure build-up. The pressure signature of the buildup period and its derivative is plotted on semi-log and log-log coordinates to develop Horner and diagnostic plots, respectively. Wellbore storage, skin and transient flow effects have been observed in the DST analysis which is an indication of the hydrocarbon bearing reservoir in the zone of interest. The value of wellbore storage effect is low which predicts the flow of liquid hydrocarbon into the well bore from the reservoir. Average pressure of the investigated zone has been estimated which is higher than the water column pressure. The higher average reservoir pressure also authenticates the presence of oil reservoir.Bangladesh J. Sci. Res. 29(1): 19-29, June-2016

The analysis of pressure buildup data in a well following a multiple-rate flow test is made more difficult if the well is near a fault. Although both these cases have been treated independently in the literature, the combination leads to a more complicated expression for the predicted pressure behaviour after shut-in. The theoretical expressions for the post shut-in pressure and the skin on the well are derived from first principles. The predicted pressure response is compared with the measured pressure buildup data from a well in the Apiay field in Colombia, and values are obtained for the permeability of 615 mD and distance to a fault of 160 m that provide a good fit to the observed, data. The distance to the fault obtained from the model matched exactly the distance to the main Apiay fault determined from :independent seismic data. Data is also presented in the form of the response surface on which the objective function minimization takes place. Introduction In most instances, drillstem testing (DST) of a well to determine productivity and reservoir properties in the vicinity of the well is completed before a thorough analysis can be made of the results. Where there are reservoir heterogeneities in the area, such as a sealing fault this often leads to a change in the slope of the usual Homer-type plot near the end of the test and it is not clear whether this increase in slope is due to the effect of the fault or is an artifact of the measurement. As several instances of this late-time increase in slope had been found in analysing drillstem test data from the Kl and K2 formations of the Apiay field in Colombia, the theoretical behaviour of a well near a fault has been analysed in order to compare this with the observed pressure behaviour in the final buildup period of a multi-rate DST The Apiay reservoir is particularly suitable for testing this type of theoretical response as it is a very clean and uniform sand of fairly high permeability with some well-defined faults and possible minor faults in the area. Theoretical Pressure Response We consider initially a DST carried out in an infinite-acting reservoir for times where wellbore storage effects have become negligible. The usual methods for treating the pressure buildup in a DST with multiple rates before shut-in are described in Earlougher(1), Matthews and Russell(2) and Lee(3). The first method, used for rates that do not vary substantially, is to calculate a production time defined by(1): Equation. (Available In Full Paper) where Vp is the total volume produced since the last pressure equalization and q is the constant flowrate just before shut-in. This is used in a Homer-type plot of PW5 as a function of Equation. (Available In Full Paper) where δt is the time since the shut-in of the well. When the flowrates before shut-in vary considerably, it is necessary to use a more elaborate multiple-rate analysis.

Summary Field data indicate that in some instances signify cant after flow occurs after a well is considered shut in. Analysis method for after flow-dominated pressure buildup data is presented whereby the PD-TD model describing the transient behavior of the well can be directly obtained by matching a log-log plot of the rate-normalized pressure vs. time data to published type curves. The PD-TD model thus obtained allows a rigorous mathematical superposition analysis to be performed on the same data with results equivalent to those obtained from the normalizedtype-curve analysis. The work demonstrates that rate normalization must be based on total after flow rates, confirming with field data Perrine's assumption that total rate should he used in multiphase flow analyses. Dramatic changes in character are seen between the rate-normalized pressure vs. time and the conventional pressure vs. time log-log data plots for low permeability stimulated wells. Several field examples demonstrate the application of this simple and effective technique. Introduction Currently, wellbore storage type curves based on the assumption of a constant wellbore storage coefficient are used to evaluate after flow-dominated data from pressure alone. For pumping wells, the ability to calculate reliable buildup pressures and corresponding after flow rates, along with the wellbore storage coefficient variation with time, has been reported in several papers. More direct bottom hole measurements of after flow and pressure using production logging tools during pressure buildup tests production logging tools during pressure buildup tests have been reported recently by Meunier et al. In the U. S., wellbore storage effects often characterize pressure buildup tests because the majority of domestic wells are produced by rod pumps where after flow dominates during produced by rod pumps where after flow dominates during the buildup. This type of completion can result in long periods of wellbore storage during test situations even for periods of wellbore storage during test situations even for stimulated wells. The first attempt to use both pressure and after flow rate data was presented by Gladfelter et al in 1955. They suggested that the pressure rise after shut-in divided by the instantaneous change in rate caused by after flow should be plotted vs. the logarithm of shut-in time. This resulted in a modified Miller-Dyes-Hutchinson (MDH)buildup plot. The validity of this approach was confirmed by Ramey" in 1965, and extended by him to include wellbore unloading effects during drawdown testing. About this same time Wine stock and Colpitts proposed a similar rate normalization of pressure for proposed a similar rate normalization of pressure for drawdown analysis of gas wells when rates were monotonically declining during drawdown tests. Rate variations were not a result of wellbore storage effects but were more a result of a nearly constant wellbore pressure test condition. Their rate normalization of the pressure test condition. Their rate normalization of the pressure data was simply an attempt to make a constant pressure data was simply an attempt to make a constant rate analysis from essentially constant wellbore pressure data. A computer study by Lee et al. basically confirmed the validity of the Wine stock and Colpitts rate normalization analysis procedure. Additional discussion of the Gladfelter et al. and the Wine stock and Colpitts normalization methods was given by Ramey in 1976. Rate-normalized type-curve plots of the Gladfelter etal. and Wine stock and Colpitts example data show that after normalization virtually all the data were on the semilog straight line. Rate normalization linearized all the data in both instances. However, tests on low-permeabilityoil and gas wells with large negative skins often cannot be analyzed using their suggested rate-normalized pressurevs. logarithm of time plotting approach because the data may not reach the semilog straight line even after normalization. A simple method of analyzing after flow-dominated pressure buildup data is presented. The PD-TD model pressure buildup data is presented. The PD-TD model describing the transient behavior of the well may be obtained directly by matching a log-log plot of the rate-normalized pressure vs. time data to published type curves. The PD-TD model thus obtained allows a rigorous mathematical superposition analysis to be performed on the same data with results equivalent to those obtained from the normalized type-curve analysis. Drawdown Rate Normalization Equations Rate normalization techniques and procedures are best illustrated by first examining their application to drawdown data. Although the nature of the rate variation of drawdown data with time is different than that of after flow rate variation, the end result is the same. Also, draw down rate variations generally last much longer than after flow rate variations. The rate normalization equation given by Wine stock and Colpitts for a gas well drawdown analysis can be written as (1) JPT p. 2211

Summary A surface shut-in drillstem test (DST) procedure as used in an offshore volatile oil reservoir. Bottomhole shut-in devices were not used because of operational difficulties and anticipated hazardous problems. The type curve and Horner plots of the bottom hole pressures (BHP's) did not conform to classical theory. This paper explains the reasons and suggests a simulation procedure for the interpretation of data. Introduction The type curve plots have a very sharp flattening bend at about 15 minutes of shut-in time, which is not typical. The tubing shut-in pressures also decline while the (BHP's) increase, this is not typical either. As a consequence of these observations, we could not accept the conventional interpretations of the buildup curves. This paper describes a new theory that explains the discrepancies. The new theory relates this unusual behavior to the combined effects of these wellbore phenomena during pressure buildup testing:wellbore storage caused by rising fluid,inertial effects caused by a rising fluid column,gas/oil phase segregation in the test string.cooling of fluid in the uppermost part of the test string by seawater, andgas going back into the oil phase in the fluid column during buildup. Background The DST procedure consisted of isolating a perforated interval by a set of packers, producing the well through a tubing (drillpipe) filled with diesel oil from several minutes to 24 hours, and finally closing the well on the surface to obtain pressure buildup data. Bottom hole shut-in devices were not used because of operational difficulties and anticipated hazardous problems during inclement weather in this hostile offshore environment. BHP's were recorded for all wells; tubing pressures and temperatures were recorded for some wells. Fig. 1 is the Homer plot of BHP and tubing pressures for a typical high-flowrate DST. Table 1 contains data needed to analyze Fig. 1. Tubing pressure and temperature are shown in Fig. 2. Fig. 3 is the type-curve plot of BHP differences vs. shut-in time. The following questions have arisen.Where is the correct straight-line portion of the Homer plot?Does deviation of the Homer plot from the early straight-line segment indicate reservoir boundary effects?Why does tubing pressure decline with shut-in time?Why does the type-curve plot have an unusual sharp bend from the early unit-slope straight-line segment? This paper provides qualitative answers to these questions based on theoretical considerations that relate to a set of complex wellbore storage phenomena. These phenomena include rising fluid wellbore storage and inertia effects at early shut-in times, gas/oil phase redistribution, wellbore fluid cooling by sea water in the top portion of the tubing, and gas dissolving back into the oil phase. Some aspects of complex wellbore storage problems and other wellbore phenomena are reported in the literature. Earlougher summarized results of the work on falling and rising wellbore storage, compressive wellbore storage, and phase redistribution wellbore effects-all before 1975. Fair's recent paper provides a promising new method of quantifying phase redistribution effect. Phase change of steam (flashing) caused by wellbore cooling for geothermal wells and its effect on well transient behavior was reported by Gringarten and Miller. JPT P. 173^

This work investigates the effects of initial reservoir pressure on the computation of effective permeabilities for homogeneous and heterogeneous solution-gas-drive reservoirs which are initially above the bubble-point pressure. The heterogeneous system considered is that of a two-zone composite reservoir with different absolute permeability and/or relative permeability curves in the two regions. It is shown that the initial effective oil permeability may be estimated from drawdown and buildup pressure data. It is also shown that one can obtain accurate estimates of computed effective permeabilities as pointwise functions of wellbore pressure from drawdown data during the time period that the wellbore pressure is less than the initial bubble-point pressure. From the estimates of effective permeabilities obtained, it is shown that for the case of homogeneous solution-gas-drive systems one can construct approximate effective (or relative) permeability curves by nonlinear regression analysis techniques. For the case of heterogeneous solution-gas-drive reservoirs it is shown that if the initial reservoir pressure is above the reservoir bubble-point pressure then the effective oil permeability of the inner or the outer zone might be obtained from the analysis of drawdown and/or buildup pressure data.

American Institute of Mining, Metallurgical and Petroleum Engineers, Inc. Discussion of this paper is invited. Three copies of any discussion should be sent to the Society of Petroleum Engineers office. Such discussion may be presented at the above meeting and, with the paper, may be considered for publication in one of the two SPE magazines. Abstract Methods of analyzing pressure buildup data from stabilized wells have been presented by several authors. All of the methods attempt to use some type of straight-line Horner plot. This paper must be considered controversial in that it shows that there is no theoretical basis for expecting a straight-line Horner plot from a stabilized well's pressure buildup data regardless of the "producing time" used. A simplified method is presented which analyzes buildup data by working with only the pressure change caused by the negative rate induced when the well is shut in. The pressure change that would have occurred had the well not been shut in is taken into account before the pressure plot is constructed. This approach permits the calculation of the undamaged transmissibility of the reservoir, the skin factor or damage ratio, and the average pressure in the well's drainage radius. These calculations can be completed with less data than required for the analysis of a pressure buildup from an infinite acting reservoir, provided the pressure change with time during stabilized production is known from past surface pressure observations, past BHP surveys, or a pressure survey just before shut-in. Introduction Except for pressure surveys run in connection with drillstem tests, most pressure buildup data are obtained from wells that are stabilized at the time of shut-in. The well known Horner type of pressure buildup analysis is based on infinite acting reservoir equations and thus is not directly applicable to most of the available pressure buildup data. Several methods have been proposed for analyzing pressure buildup data from stabilized wells, but all of the methods present some difficulties in application. The published methods attempt to adapt the Horner plot to stabilized wells. They also require trial-and-error solutions and/or reservoir data which are not normally available. The method developed in this paper does not require a trial-and-error solution nor does it require any more reservoir data than does the Horner type of analysis of an infinite acting reservoir. The Horner pressure buildup equation can be written in two terms in practical units as, The first term represents the pressure change that would occur in (tp + delta t) days, if the well had not been shut in. The second term represents the pressure increase which occurs due to introducing a -q rate at the well when the well is shut in for delta t days. van Everdingen and Hurst showed that these pressure changes are a function of the log of the time only as long as the reservoir is infinite acting. This paper presents a method of analysis that concerns itself with the second term of this equation and for a delta t small enough for the effect to be infinite acting.

This paper presents new procedures for analyzing pressure buildup data obtained from drillstem tests. The new methods apply for cases where the produced fluid does not reach the surface during the flowing period so that, the flow period represents a slug test. The combined effects of variable flow rate, short producing time and changing wellbore storage generally make it difficult to apply conventional analysis methods to DST buildup data. Two new straight-line methods for analyzing buildup data presented in this work account for both the variable rate during the flow period and producing time effects. If the buildup well-bore storage-skin group Csd exp(2s) is small, then shortly after shut-in, a well-defined straight line is obtained for both methods. Using the slope of the straight line obtained by either method, it is shown that the flow capacity (kh), the skin factor (s) and the initial reservoir pressure (pi) can be determined. If the group Csd exp(2s) is large, longer shut-in times are needed before the proper straight line can be obtained. However, for the latter case, it is shown that a multi-rate equivalent time can be constructed so that standard type-curve matching can be performed to obtain estimates of the flow capacity and the skin factor. A field example is presented to illustrate the applications of the proposed methods.

The wellbore acts as a storage tank during drawdown and build-up testing and causes the sand-face flow rate to approach the constant surface flow rate as a function of time. This effect is compounded if non-Darcy flow (turbulent flow) exists near a gas wellbore. Non-Darcy flow can be interpreted as a flow-rate dependent skin effect. A method for determining the non-Darcy flow constant using this concept and the usual skin effect equation is described. Field tests of this method have identified several cases where non-Darcy flow was severe enough that gas wells in a fractured region appeared to be moderately damaged. The combination of wellbore storage and non-Darcy flow can result in erroneous estimates of formation flow capacity for short-time gas well tests. Fortunately, the presence of the wellbore storage effect permits a new analysis which can provide a reasonable estimate of formation flow capacity and the non-Darcy flow constant from a single short-time test. The basis of the Gladfelter, Tracy and Wilsey correction for wellbore storage in pressure build-up was investigated. Results led to extension of the method to drawdown testing. If non-Darcy flow is not important, the method can be used to correct short-time gas well drawdown or build-up data. A method for estimation of the duration of wellbore storage effects was developed. INTRODUCTION In 1953, van Everdingen1 and Hurst2 generalized results published in their previous paper3 concerning wellbore storage effects to include a "skin effect", or a region of altered permeability adjacent to the wellbore. Later, Gladfelter, Tracy and Wilsey4 presented a method for correcting observed oilwell pressure build-up data for wellbore storage in the presence of a skin effect The method depended upon measuring the change in the fluid storage in the well bore by measuring the rise in liquid level. To the author's knowledge, application of the Gladfelter, Tracy and Wilsey storage correction to gas-well build-up has not been discussed in the literature. It is, however, a rather obvious application. Gas storage in the wellbore is a compressibility effect and can be estimated easily from the measured wellbore pressure as a function of time. Several approaches to the wellbore storage problem have been suggested. As summarized by Matthews,5 it is possible to minimize annulus storage volume by using a packer, and to obtain a near sand-face shut-in by use of down-hole tubing plug devices. Matthews5 and Perrine6 have suggested criteria for determining the time when storage effects become negligible. In 1962, Swift and Kiel7 presented a method for determination of the effect of non-Darcy flow (often called turbulent flow) upon gas-well behavior. This paper provided a theoretical basis for peculiar gas-well behavior described previously by Smith8. Recently, Carter, Miller and Riley9 observed disagreement among flow capacity kgh data determined from gas-well drawdown tests conducted at different flow rates for short periods of time (less than six hours flowing time). In the original preprint of their paper, Carter et al.9 proposed that the discrepancy in flow capacity was possibly a result of wellbore storage effects, Results of an analytical study of unloading of the wellbore and nonDarcy flow were recorded by Carter.10 In the final text of their paper, Carter et al.9 stated that they no longer believed wellbore storage was the reason for discrepancy in their kgh estimates. In view of the preceding, this study was performed to establish the importance of non-Darcy flow and wellbore storage for gas-well testing. In the course of the study, a reinspection of the previous work by van Everdingen1 and Hurst2 was made, and the basis for the Gladfelter, Tracy and Wilsey4 wellbore storage correction was investigated and extended to flow testing. WELLBORE STORAGE THEORY As has been shown by Aronofsky and Jenkins,11,12 Matthews,5 and others, flow of gas can often be approximated by an equivalent liquid flow system. The following development will use liquid flow nomenclature to simplify the presentation. Application to gas-well cases will be illustrated later. First, we will use the van Everdingen-Hurst3 treatment of wellbore storage in transient flow to establish (1) the duration of wellbore storage effects, and ( 2) a method to correct flow data for wellbore storage.

Summary Methods for analyzing buildup data following a short flow period are presented, discussed, and illustrated. A new type curve for uniform-flux and infinite-conductivity vertically fractured wells is presented. By matching buildup data with this new type presented. By matching buildup data with this new type curve, we can determine the dimensionless flowing time before shut-in. A method for converting buildup data to equivalent drawdown data is discussed. This method can be used to combine buildup and drawdown data to obtain a longer band of data for type-curve matching. This method canbe used for constant-rate production, constant-pressure production, and for the case where both pressure and rate production, and for the case where both pressure and rate vary during production. Introduction Over the past decade, the use of type-curve matching toanalyze pressure data has gained increasing acceptance. The advantages and dis advantages of type-curve matchingare well recognized and the procedure has become astandard tool to analyze data qualitatively, to identify flow regimes, and, quantitatively, to determine iformation parameters. Virtually all type curves available in the literature examine the pressure response at a flowing well-i.e., only drawdown solutions have been examined. These type curves may be used to analyze shut-in pressure behavior provided that the flowing time before shut-in is provided that the flowing time before shut-in is significantly longer than the maximum shut-in time. This limitation has prevented type-curve analysis ofpressure buildup data following short flow periods. pressure buildup data following short flow periods. The effect of a short flow period on pressure buildup data influenced by either wellbore storage or vertical fractures was presented recently. It was shown that significant errors result if proper care is not taken in analyzing pressure buildup data when the producing timeis short. Procedures to account for the influence of producing time were outlined. producing time were outlined. The objectives of this paper areto present anew correlation that considerably simplifies the use of buildup type curves for vertically-fractured wells givenin Ref. 1, andto suggest a procedure to analyze pressure buildup data by means of drawdown type pressure buildup data by means of drawdown type curves. The procedures discussed here can be applied to single- or multiwell tests, to data obtained after shortor long flow periods, to constant or variable flow rates, and to virtually all wellbore conditions including data influenced by wellbore storage and skin and fracturesof finite or infinite conductivity. Theory The basic pressure buildup equation based on the principle of super position is given by principle of super position is given by (1) Here, t is flowing time and Delta t is shut-in time. Thesymbols p and t denote dimension less wellbore pressure drop and dimension less time, respectively, and are defined as follows. JPT P. 904

Shunnan Block in North-West China is one of the toughest HPHT gas reservoirs with vertical depth over 7500 m, formation temperature over 200 and pressure gradient varying from 1.3 to nearly 2. The condition is close to temperature and pressure limit of well testing tools, therefore, the tools are hard to be sent to downhole and chances are that well testing operations usually failed. The pressure buildup data are with bad quality and needed to be converted into downhole data. Meanwhile, it's hard to diagnose accurate flow regimes and interprete because the block is typically carbonate reservoirs with porous medium including pores, natural fractures and caves. In this paper, we reviewed the exploration wells in this block and find that interpretation by pressure buildup or transient production data can only reflect part of the formation information; therefore the two kinds of data are combined to get more accurate interpretation results. For pressure buildup interpretation, three models including dual porosity model, composite model, and dual porosity with composite model are chosen and compared. For the production data, dual porosity model with boundary is selected because the wells usually show characteristics of multiple porous medium and boundary dominated flow. Parameters interpreted from pressure buildup data are simultaneously transferred into the model for production data. Results show that the combined interpretation by pressure buildup and production data can reduce the un-uniqueness of models as well as enhancing the accuracy of formation and wellbore parameters evaluation. The model and parameters can satisfy both pressure buildup and production data history. Although Shunnan block is considered as one greatly promising reservoir from the short period open flowing, the combined interpretations suggest very limited drainage volume. Reasons for this paradox phenomenon may be that the wells are severe contaminated by drilling fluid, or the wells were only producing gas in caves and natural fractures nearby the wells while other caves are not connected.

Fluid flow rate and density at the sandface measured during drawdown in a drillstem test (DST) are key measurements that are normally unavailable using conventional techniques. Production logging flowmeters cannot be used in most cases, and the presence of wireline in the test string when the well is flowing is sometimes an undesirable complication. Furthermore, intrusive devices such as spinners have proven especially vunerable to damage by cuttings and produced sand. This paper presents field data recorded using a prototype downhole multiphase flowmeter sensor that is particularly well-adapted to DST. This fullbore tool can be positioned below the test valve and provides continuous mass flow rate, density, pressure and temperature measurements. Drawdown data were interpreted to yield permeability, skin and reservoir limits which were comparable to the values determined from a conventional buildup interpretation. In addition, it was possible to monitor any changes in skin while flowing the well during the cleanup period.