SPE 164871 Well-head Pressure Transient Analysis Charidimos E. Spyrou, SPE, Schlumberger, Peyman R. Nurafza, SPE,E.ON E&P, Alain C. Gringarten, SPE, Imperial College Copyright 2013, Society of Petroleum Engineers This paper was prepared for presentation at the EAGE Annual Conference & Exhibition incorporating SPE Europec held in London, United Kingdom, 1013 J une 2013. This paper was selected for presentation by an SPE program committee following review of information contained in an abstract submitted by the author(s). Contents of the paper have not been reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflect any position of the Society of Petroleum Engineers, its officers, or members. Electronic reproduction, distribution, or storage of any part of this paper without the written consent of the Society of Petroleum Engineers is prohibited. Permission to reproduce in print is restricted to an abstract of not more than 300 words; illustrations may not be copied. The abstract must contain conspicuous acknowledgment of SPE copyright. Abstract Pressure Transient Analysis (PTA) of bottom-hole pressure data (BHP) is a well-established method for estimating reservoir flowparametersandidentifywellbehaviour.Unfortunately,permanentrecordingofbottom-holedataisnotalways operationally possible, for example in the case of high pressure and high temperature (HP/HT) reservoirs. On the other hand, most wells are equipped with gauges at the well head which record well head pressure data (WHP) continuously. This paper investigates the feasibility of using WHP data for identifying well test behaviour. The objective is to assess the ability to derive key well and reservoir parameters from WHP data in the absence of BHP data, focusing primarily on the estimation of permeability and skin. Three different actual reservoir fluid and wellbore conditions were studied: a water injector with a single phase fluid in the wellbore and a reasonably constant fluid density; a dry gas well, also with single phase in the wellbore but changing fluid density; and a gas condensate well with multiphase flow and varying density in the wellbore. In each case, both WHP and BHP data were available. These WHP and BHP data were analysed separately with conventional PTA methods in order to compare resulting permeabilities and skin factors. WHP data were then converted to BHP data, using methods available in the literature in the water and dry gas cases, and two different approaches developed in this work for the gas condensate case.In the case of the water injector, analysis of both the original WHP data and the converted BHP data provides a good estimate of permeability while overestimating the skin factor. A correction can be applied to the WHP derived skin, to match the BHP skin value.In the dry gas and gas condensate cases, WHP analysis overestimates both permeability and skin factor. In the dry gas case, the WHP derived permeability and skin can be corrected to yield the BHP values, whereas PTA on the converted BHPdata doesprovidethe correctpermeability.Finally,inthegascondensatecase,itispossibletoobtainthecorrect permeability from converted BHP in the absence of phase redistribution, whereas the skin factor remains overestimated. Introduction One of the key tasks of reservoir engineers is to identify well and reservoir behaviour, especially during the appraisal phase of a field development. Bottom-hole (BH) Pressure Transient Analysis (PTA) is a well-established method for estimating reservoirflowparametersandidentifyingwellbehaviour.Permanentrecordingofbottom-holedataisnotalways operationally possible, however, for example in the case of high pressure and high temperature (HP/HT) reservoirs. On the otherhand,wellhead(WH)gaugesarepresentonmostwellsandWHpressuresareoftencontinuouslyrecordedby operating companies. The availability of WHP data raises the question of whether this data can be used in the absence of BH pressures to identify well and reservoir behaviour. The aim of this work is to investigate the ability of WH data to provide useful estimations of key reservoir parameters such as permeability and skin. There are several advantages to be able to derive useful information from WH data. The cost of recording WH data (which are recorded anyway in most cases) is much less than that of a downhole survey, and the risks associated with running tools in the wellbore are eliminated. This would be particularly useful in HP/HT conditions where wells cannot be tested due to harsh downhole conditions, such as high temperatures and pressures, completion integrity issues and tubing restrictions.Despite its significance, this area has not yet been fully explored. Smith (1950) was the first to propose a WH to BH conversion algorithm for dry gas wells in flowing conditions. Cullender and Smith (1956) developed a procedure which is widely used to calculate bottom-hole pressures in gas wells and makes no assumptions for temperature and compressibility. SeveralmethodswerealsodevelopedtoaccountforthepresenceofliquidinthewellboresuchasthatofGovierand Fogarasi (1975) and the modified Cullender and Smith equation by Peffer et al. (1988). All the above methods provide 2SPE 164871 satisfactoryresultsbutarelimitedtoflowingconditions.DallOlioandVignati(1998)werethefirsttodevelopa methodology which allows the use of WH pressure data for test interpretation purposes, by correcting the analysis results to reservoir conditions. Fair et al. (2002) presented a methodology to categorize wells based on fluid type and reservoir and wellbore behavior, and a procedure to test wells from the surface. This paper assesses the ability of PTA of WH pressure build ups (PBU) to identify well behaviour. Three different cases have been investigated according to the fluid behaviour in the wellbore: a water injector case where there is a single phase fluid in the wellbore and the density is reasonably constant; a dry gas case, still single phase but with with density variations in the wellbore; and a gas condensate case where a two-phase fluid is present in the tubing and the density varies. The study is primarily focused on determining the impact on permeability and skin. The analysis of each three cases will be presented individually. The approach used was to study WH datasets for which corresponding BH pressures were available and to compare WHP and BHP analysis results. The comparison was then extended to the analysis of BHP converted to bottomhole conditions. Methodology, Analysis and Discussion Case 1: Water Injector Log-log plot comparison WHP and BHP for a water injection well are shown in Figure 1. There is only a single-phase fluid in the wellbore, with a constant density. The BHP and WHP derivative curves in the log-log plots of Figure 2 are very similar, but the corresponding pressure change curves are different, with the WH curves being above the BH curves. This implies that WHP data should yield the same permeability as BHP data, but would significantly overestimate the skin. Figure 1: Pressure and Rate history. SPE 1648713 1101001000100000.0001 0.001 0.01 0.1 1Pressurechangeandderivative(psi)Time(hr)WH1101001000100000.0001 0.001 0.01 0.1 1Pressurechangeandderivative(psi)Time(hr)BH PTAPTA of the data (Figure 3) confirms the observations as shown in Table 1. Fall-off 6 was the only period long enough to reach radial flow stabilization. BHP and WHP analyses yield the same permeability value, which was then used to calculate the skin factor from the other fall offs. As expected from the log-log plots, WHP data overestimate the skin effect. The skin calculated from BHP data decreases with successive fall-offs, suggesting that the well is getting stimulated (possibly through the creation of a fracture). Such a trend is not obvious with WHP, as skin variation is more random.

Analytical Approach The above observations can be verified analytically. Fig. 4 presents the difference between BH and the WH pressures vs. time for Fall Off 1 (FO1). The difference is higher during the first few seconds (which means that the WH pressure is falling off faster than the BH) then stabilizes, indicating that BHP is the sum of WHP and a constantop, the weight of the water column during that fall off:pw] = pwh +op. The BHP, pwf, during a fall off after radial flow has been reached is given by Eq. 1: BHWH FO1 k(mD)2727 Skin37.593.5 FO2 k(mD)2727 Skin26.770 FO3 k(mD)2727 Skin22.752.9 FO4 k(mD)2727 Skin20.892 FO5 k(mD)2727 Skin20.465.2 FO6 k(mD)2727 Skin313.9 Table 1: Permeability and skin estimations from BH and WH data. Figure 2: Log-log plots comparison between BH and WH data.Figure 3: PTA on BH and WH (FO6). BH dataWH dataPressure change and derivative(psi) Pressure change and derivative(psi) Pressure change and derivative(psi) Pressure change andderivative(psi)Pressure change and derivative(psi) Pressure change and derivative(psi) Time (hr)Time (hr) Time (hr) Time (hr)Time (hr) Time (hr) FO1FO2 FO4 FO5 FO3 FO6 4SPE 164871 1400190024002900340039000 0.2 0.4 0.6 0.8p(psi)Time(hr)BHWH020004000600080001000026.4 26.6 26.8 27 27.2Pressure(psi)Time(hr)MeasuredBHEstimatedBHWH101001000100000.001 0.01 0.1 1Pressurechangandderivative(psi)Time(hr)0204060801000 20 40 60 80 100WHskinfactorBHskinfactorWaterInjectorWHVsBHCorrectedVsBHidealp -pw] =162.6qBkh[logt +logkqctw2 -S.2S +u.87s.........(1) The WHP, pwh, is therefore given by Eq. 2: p -pwh = op +162.6qBkh[logt +logkqctw2 -S.2S +u.87s =162.6qBkhjlogt +logkqctw2 -S.2S +[u.87s +kh162.6qBop[(2) Analysis of WH data yields a higher total skin factor than analysis of BH data, but the same permeability as the derivatives of the LHS of Eqs 1 and 2 with respect to logt are identical. The same results are obtained with actual and converted BHP data for FO1 are shown on the Cartesian plot of Fig.5 and Fig.6 respectively. The match is very good in Fig. 5 except at early times for the same reason as in Fig. 4. The high initial converted BH pressures yields a higher converted pressure change curve in Fig. 6 and therefore a higher skin effect than the actual BH pressure data. The permeability is the same. It is possible to use the permeability derived from the WH analysis to calculate the skin. Since the high WH pressure at the first seconds of the fall-off is a result of the water injection, the pressure values corresponding to those time steps should be eliminated. Then the converted BH pressures can be used to estimate skin by rearranging the semi-log radial flow approximation equation (Eq. 3). This correction method can yield satisfactory results as it can be seen in Table 2. p(t) =12(lnt +u.8u9u7 +2s).......(3) Figure 4: Subtraction of WH from BH pressures (FO1). Figure 5: Plot of the estimated along with the actual and the WH data (FO1). Figure 6: Log-log comparison between estimated and actual BH pressures (FO1). Figure 7: WH and corrected skin factor Vs BH. SPE 1648715 1600200024002800Pressure(psia)x1x2x3x0 100 200 300 400 500 600 700GasRate(MMscf)Time(hr)s =2pD-IntD-0.809072...(4) wherep =khAp141.2qBwand t =0.000264kAtqctw2 It is shown in Fig.7 that the correction provides satisfactory results of the skin. The WH vs. BH curve however displays an inconsistency between FO3 and FO4. After FO3 the injection rate was increased. Since the WH pressure is affected by the injection of the water it can be assumed that increased injection rate will result to a higher WH pressure at the beginningofthefall-off.ConsequentlyahigherApcurveoftheWHdataisexpectedandthereforeanincreasedskin estimation. Case 2: Dry Gas For the purpose of this study and as a limit to the methods described, a dry gas well is considered. This is a well with an oil production of less than 10bbl/MMcf (Fair et al. (2002)). The complexity of this case is greater than the water injection case. In the wellbore there is still single-phase but now the density varies along the wellbore due to the compressibility of gas. Figure 8: Rate and pressure history. Log-log plot comparison and WTA ForthiscaseonlytwodatasetsofBHwiththeircorrespondingWHpressurewereavailable.ThemajorityoftheWH pressures at each time step were interpolated and only a few values of pressure were measured. For this reason the data were deconvolved to generate drawdown responses and compare the log-log plots. As shown in Figure the Ap and derivative curves have very similar shapes and the WH curves seem to be slightly shifted downwards. Consequently we expect the WH datatopredicthigherpermeabilitiesandskinfactors.ThisisconfirmedbyPTA(Fig.10)whichisconsistentwiththe observations. WH data slightly overestimated permeability and skin for both cases (Table 3). WHBHCorrected FO1 Skin 93.737.536.3 FO27026.725.3 FO352.922.722.6 FO49220.822.4 FO565.220.421.8 FO613.931.8 Table 2: Comparison of the corrected skin against WH and BH estimations. 6SPE 164871 1.E+051.E+061.E+071.E+081.E+090.001 0.01 0.1 1 10 100Pressurechangeandderivative(psi)Time(hr)WH1.E+051.E+061.E+071.E+081.E+090.001 0.01 0.1 1 10 100Pressurechangeandderivative(psi)Time(hr)BH Analytical Approach Smith,R.V.(1950)wasthefirstonetofindarelationbetweenwellhead(pwh)andbottom-hole(pw])pressuresby integrating the energy balance equation along a straight line assuming a constant tubing internal diameter and negligible variation of zI. Neglecting acceleration losses the correlation between WH and BH pressure should be represented in the form (Eq. 5): pwh2= pw]2Ks -K]......(5) Ks = c-S.........(6) S =0.0683ygL(z1)cg....(7) where Ks and K] represent the gravity forces and the friction losses respectively. Since this study focuses on Pressure Build-ups where there is no flow in the wellbore it is safe to assume that friction losses are minimum and therefore neglect K]. Eq. 5 is then used to calculate BH pressures in the form of Eq. 8. The results of the converted BH pressure are shown in Fig. 11. The error in each time step is less than 2%. BHWH BU1 k(mD)2232 Skin0.92.8 BU2 k(mD)2128 Skin-1.1-0.7 Table 3: PTA results.Figure 9: Dry gas log-log plot comparison of BH and WH data.Figure 10: PTA on WH and BH data. 1.E+061.E+071.E+081.E+090.01 0.1 1 10 100Pressurechangeandderivative(psi)Time(hr)BU21.E+061.E+071.E+081.E+090.01 0.1 1 10 100Pressurechangeandderivative(psi)Time(hr)BU1BH dataWH dataSPE 1648717 As it can be seen in the log-log plots (Fig.12), the estimated BH derivative overlays the derivative of the actual BH data that suggests that the converted data can estimate the same permeability as the actual. The Ap curve though is shifted upwardswhichindicatesthattheskinestimationwouldbegreaterthantheactual.Thisislikelytobebecauseinboth examples Eq. 8 at early times tends to underestimate the BH pressures with an error that is greater than middle and late times where the error stabilises at a lower value. The error stabilisation explains why the derivatives are the same and the higher error at the beginning explains the higher Ap curve. pw] = pwh_1Ks....(8) DallOlio and Vignati (1998) in their paper suggest that the value of the permeability derived by the interpretation of the WH data can be corrected to match the value that a proper BH interpretation would yield. Using Darcys law for single phasegas(Eq.9)andSmithsformula(Eq.5)theyfoundacorrelationbetweenthecorrectpermeabilityandtheone estimated by WH data (Eq. 10). qg = 7.uS 10-4kgh[pr2-pw]2z1jIn[rcrw-0.75+st+qg[....(9) k =z1(z1)rc]kc]Ks ..(10) where the subscript ref is referring to the values that were used for the WH interpretation and Ks refers to the gravity losses of Eq. 5. For viscosity, z factor and Temperature without subscript, values that represent reservoir condition should be used. Similar to what was done for the water injector case, by knowing the corrected permeability and with an estimation of the BH pressures the skin can be estimated using Eq. 4. The results are in a very good agreement with the actual as it can be seen in Table 4 as well as Fig. 14 and Fig. 15. Figure 11: Dry gas converted BH pressures against actual BH and WH pressures.Figure 12: Log-log plots of the estimated BH against actual BH and WH pressures.1.E+061.E+071.E+081.E+090.01 0.1 1 10 100Pressurechangeandderivative(psi)Time(hr)BU21.E+061.E+071.E+081.E+090.01 0.1 1 10 100Pressurechangeandderivative(psi)Time(hr)BU1BH dataWH data Estimation8SPE 164871

Case 3: Gas Condensate When a gas condensate reservoir pressure drops below the dew point pressure, liquid condensate is formed. This leads to the presence of a two-phase fluid in the wellbore during production. Due to the compressibility of the gas and condensate, the densityvariesalongthewellbore.Inadditiontothatthehold-updepthisnotconstantandduringtheshut-in,liquid reinjection in the reservoir may take place.The exhibition of this complex behaviour makes the study of this case more difficult than the previous two.For this rich gas condensate reservoir two examples of build-ups have been studied (Figure 1). For the first one the reservoir pressure is above the dew point pressure whereas for the second the pressure is below and a condensate bank is formed. BHWHCorrectedPBU1 k(mD)223224 Skin0.92.80.6 PBU2 k(mD)212822 Skin-1.1-0.7-1.4 Table4:Tableofthecorrectedkandskinusing Dall'Olio and Vignati method. Figure 13: Rate-normalised log-log plot of the two build-ups.Figure 14: WH and corrected permeability values Vs BH. Figure 15: WH and corrected skin factor values Vs BH.Figure 16: Rate and pressure history. 1.E+061.E+071.E+081.E+090.01 0.1 1 10 100Normalizedpressuerechangeandderivative(psi)Time(hr)BH dataWH data50015002500350045005500x1x2x3x4x5x0 100 200 300 400Pressure(psi)Production(MMscf)Time(hr)BU2RatesPressures72007500780081008400x1x2x3x4x5x0 100 200 300 400Pressure(psi)Production(MMscf/d)Time(hr)BU1RatesPressuresSPE 1648719 Log-log plot Comparison and WTA In contrast to what it was observed in the water injector and dry gas case, the study of the log-log plots is not helpful as the plots seems not to display any specific trends (Figure 1 17). Despite that, it is expected that the WH interpretation would overestimatedpermeability.ItisnotclearwhattheestimationoftheskinwouldbebecauseinFig.17LHS,WH interpretationislikelytooverestimateitwhileinFig.17RHS,WHcurvesareshifteddownwardsandtheWHskin estimation would be the same as the BH skin estimation. The WTA on the two datasets (Fig. 18) confirms the observations and the results are shown in Table 5. PTA on the BH pressures returned different permeabilities for the two build-ups. Reservoir pressure dropped below the dew point pressure at some time between the two build-ups. A condensate bank therefore should exist around the well at the time of the second build-up. Consequently the BU2 permeability represents the permeability of the condensate bank. If the shut-in period were longer an increased in mobility would have been seen as a second stabilization of the derivative at the BU1permeability,whichrepresentsthereservoirpermeability(Fig.19).TheBU1skinvaluesinTable5representthe wellbore skin effect, whereas the BU2 skin values corresponds to the total skin factor which is the sum of wellbore and condensate bank skins. BHWH BU1 k(mD)7.4613.7 Skin-1.493.79 BU2 k(mD)1.713.59 Skin31.734.6 Table 5: PTA results for the Gas Condensate case.Second stabilization (outside condensate bank zone) First stabilization due to condensate bank Figure 17: Gas Condensate log-log plot comparison BH and WH.Figure 18: WTA on Gas Condensate data. Figure 19: Gas Condensate pressure and deri vati ve behaviour1.E+061.E+071.E+080.001 0.01 0.1 1 10Pressurechangeandderivative(psi)Time(hr)BU11.E+061.E+071.E+081.E+090.01 0.1 1 10 100Pressurechangeandderivative(psi)Time(hr)BU2BH dataWH data1.E+051.E+061.E+071.E+080.001 0.01 0.1 1 10Pressurechangeandderivative(psi)Time(hr)BH1.E+051.E+061.E+071.E+081.E+090.001 0.01 0.1 1 10Pressurechangeandderivative(psi)Time(hr)WH10SPE 164871 Analytical Approach As a first step an attempt was made to correct the WH estimations of permeability and skin as was done for the water injector and dry gas cases. Results were less representative of the actual values and even of the WH derived parameters.When flow is multi-phase in the wellbore and the reservoir is shut-in, liquid falls back and reinjection may occur. Due to the difference in the density between the two phases Wellbore Phase Redistribution (WPR) takes place (Ali et al. 2005).Thedenserphasemovestothebottomofthewellwhereasthelighterphaserisestothesurface.Becauseof compressibility effects, WPR results to an increase in the wellbore pressure which is dissipated through the formation until equilibrium is reached between the reservoir and the wellbore (Ali et al. 2005).After WPR is over, the well exhibits a segregated phase distribution (Nurafza et al. 2009) where the gas column lies above the oil column. Because of that two different pressure gradients are observed, one for the gas column and one for the oil column. As a result there is no direct pressure communication between the WH and the reservoir (Fair 2001, Fair et al. 2002). The pressure communication can only be established when and if all of the liquid is reinjected in the reservoir and there is only single-phase gas present in the wellbore (Fair et al. 2002).Since there is no pressure communication between the WH and the reservoir WH derived parameters cannot be correctedtomatchtheactual.TheonlywayforwardistoconvertWHtoBHpressures.Inthefollowingsectionstwo conversion methods will be presented and discussed. WH to BH ConversionAn operational WH to BH pressure algorithm should be able to take into account the temperature profile in the wellbore through time (Fair et al. 2002, Hasan et al. 2005). In HP/HT fields the WH temperature can go up to 300oF due to the flow of the fluids from the reservoir. During shut-in the flow stops allowing for the WH to cool down. It is well known that this phenomenoncould resultsinadrop of WHP during build-uplatetimes.In Fig.21areal fieldexampledisplayingthis behaviour can be seen. In addition to the wellbore-temperature profile, a conversion algorithm is necessary to account for the change in fluid properties for different pressures and temperatures. Consequently a PVT model should be created and used. 1st Method: Modified Peffer et al.(1988) Equation In their paper Peffer et al. (1988) developed a method to calculate BH pressures using a modification of Cullender and Smith (1956) equation (Eq. 11) to account for the presence of liquid in the gas to obtain satisfactory results for flowing conditions. Figure 20: Rate-normalized plot of the two built-ups.Figure 21: Example of WH pressure decreasing during a shut-in in a Gas Condensate well. 1.E+061.E+071.E+081.E+090.001 0.01 0.1 1 10 100Normalizedpressurechangeandderivative(psi)Time(hr)BH dataWH dataSPE 16487111 ] _pTZ dp66]Mq2dS[LD+[ pTZ2 _ =yg53.34pw]pr]

...(11) Theaboveformoftheequationisonlyapplicabletodry-gaswells.TheadjustmentthatPefferetal.(1988) proposed, so that the presence of liquid would be accounted, was to change the surface gas gravity in Eq. 11 with a wet-gas specific gravity that can be calculate by Eq. 12 developed by Rzasa and Katz (1945). When the molecular weight of the condensate is not known it can be estimated by Eq. 13 (Cragoe 1929). ywg =yg+4,S84ycRg1+132,800ycMcRg.....(12) Ho =44.29yc1.03-yc......(13) Since a shut-in is investigated, the methodology used in this study was to consider friction losses equal to zero and therefore neglect the friction term from the Eq. 11 and apply the Peffer et al. (1988) modification. The conversion equation for a shut-in then takes the following form: ]1zp Jp =ywg53.34pw]pr]

...(14) The equation was tested and validated with two dataset where both the WH and the corresponding BH pressures were known. To implement the equation a Visual Basic Application (VBA) Macro was developed and an existing PVT model was used. The Macro was designed to run an algorithm that divides the wellbore in 100ft segments and take the WH pressure and Temperature as initial inputs. It then calls for the PVT model to calculate z, pg, po and Ho for that pressure and temperature. These parameters are used to estimate wet gas specific gravity with Eq. 12. The next segments pressure is then calculated by Eq. 14 which is implemented with the trapezoidal rule. The procedure is repeated until the depth of the bottom-hole gauge is reached.The algorithm was found to be very sensitive to temperature and produced erroneous results when temperature was changing with time. For this reason in the simulations the temperature profile was varied versus the well depth but not versus time. The results of the conversion can be seen in Fig. 22 where the estimated BH pressures are plotted against the actual. The error in each time step is less 2%. The log-log plots of the two datasets were then compared (Fig. 23). The estimated derivative is similar to the actual and in the case of BU2 seems to overlay it. The Ap curve though is much higher for the estimated pressures. This is because the method under predicts the pressure at early times whereas at middle and late times it over predicts it. The derivative seems to be the same because the error between estimated and actual values is stabilizing after early times. Results indicate that the converted BH pressures might estimate the permeability correctly but overestimate skin. This is confirmed by the PTA and the results are displayed in Table 6 as well as in Fig. 24 and Fig. 25. Figure 22: Estimated BH pressures against the actual (modified Peffer method).190029003900490059000 1 2 3 4 5Pressure(psi)Time(hr)MeasuredBHEstimatedBH73007500770079000 1 2 3 4 5Pressure(psi)Time(hr)MeasuredBHEstimatedBH12SPE 164871 2nd Method: Adding Column Weight A simplified method was also tested. The idea was to use the PVT model to find an average density of the fluid for each 100ft segment since an equation that gives wet gas specific gravity is known (Eq. 12). The density was used to find the weight of the fluid column and add it to the WH pressure. The procedure is repeated until the bottom-hole gauge depth is reached. Temperature was varied versus time in agreement with a WH temperature profile that was recorded during a shut-in performed on another well in the same field. The results of the estimations obtained by the adding column weights are plotted in Fig. 26. The method provides very good results for the calculation of pressure at each time step. The error, although high in the first few seconds, is less the 2% for the rest of the built-up. The log-log plots of the estimated data (Fig. 27) though, are different from the actual BH pressure log-log plots. This is because the error is changing at each time step and is not reaching a stabilization point as happened in the 1st method. Therefore the results are not suitable for a PTA as they yield incorrect results. Nevertheless the method can provide a good estimation of the pressures. The increasing error between estimated and actual data though is an indication that as the build-up progress results will be less representative. BH WH EstimationBU1 k(mD)7.4613.76.75 Skin-1.493.790.312 BU2 k(mD)1.713.591.71 Skin31.734.643.3 Table 6: Results of the WTA on estimated BH, actual and WH pressures. Figure 23: Log-log plot comparison. Estimated BH pressures against actual and WH (modified Peffer method). Figure 25: WH and corrected skin factor Vs BH. Figure 24: WH and corrected permeability Vs BH.1.E+061.E+071.E+080.001 0.01 0.1 1 10Pressurechangeandderivative(psi)Time(hr)BU11.E+061.E+071.E+081.E+090.01 0.1 1 10 100Pressurechangeandderivative(psi)Time(hr)BU2BH dataWH data EstimationSPE 16487113 Conclusions Water Injector ItisshownthatanalysisofWHdatacanyieldthesamederivativewiththeBHdata,byintegratinglog-log observations, PTA results and the analytical approach. The BH pressure during build-up can be reasonably estimated by adding the weight of the water column to the WH pressure data. The derivative curve generated by the estimated pressures matches the derivative generated by the actual data indicating that the permeability can be predicted correctly. Skin can be corrected using the permeability derived from the WH interpretation, the estimated BH pressures and the semi-log radial flow approximation, to satisfactorily match the results from BH data. Dry Gas WH data can be converted to BH using Eq. 4, with an error of less than 2%. The converted BH data yield the same pressure derivative as the actual BH data, indicating that the WH data can predict permeability quite accurately in the cases of a dry gas fluid in the wellbore. WH derived permeability can be corrected to the actual value using DallOlio and Vignati (1988) correlation. SkincanbeadjustedtomatchtheestimationofBHdata,usingthecorrectedpermeability,theestimatedBH pressures and the semi-log radial flow approximation (Eq. 4). Gas/Condensate Two methods were developed to calculate BH pressures providing reasonable results. The error between estimated andactualpressuresinthemodifiedPefferetal.(1988)methodisstabilizingatmiddleandlatetimesand consequently leads to a derivative that is very similar with the one from BH data. PTA indicates that permeability can be estimated using the converted pressures. Skin factor though is overestimated. A good estimation of the pressures can be provided by using the Adding Column Weight method, but the results should be used only as a sense of BH pressures, not for PTA. Figure 26: Comparison of the estimated with the actual BH pressures (adding column weights method). Figure 27: Converted data against actual and WH log-log plot comparison.190029003900490059000 1 2 3 4 5Pressure(psi)Time(hr)BU2MeasuredBHEstimatedBH73007500770079000 1 2 3 4 5Pressure(psi)Time(hr)BU1MeasuredBHEstimatedBH1.E+061.E+071.E+080.001 0.01 0.1 1 10Pressurechangeandderivative(psi)Time(hr)BU11.E+061.E+071.E+081.E+090.01 0.1 1 10 100Pressurechangeandderivative(psi)Time(hr)BU2BH dataWH data Estimation14SPE 164871 Acknowledgements ThisstudywasconductedbyCharidimosSpyrouatImperialCollegeinpartialfulfillmentofpost-graduatestudy requirements. The authors would like to thank E.ON E&P UK and Total E&P UK as well as BG Group, Carrizo, Centrica, Chevron,DanaPetroleum(E&P)Limited,DyasE&P,ENIUKLimited,ExxonMobil,Noreco,PremierandSummit Petroleum, for providing data and their permission to present and publish this material. Our appreciation goes to Basil Al-Shamma, Paul Arkley and Helene Nicole for taking time from their busy schedule to discuss and provide their ideas. Lekan Aluko is also acknowledged for his contribution. Nomenclature BFormation volume factor cCompressibility (psi-1) ygGas specific gravity yoOil specific gravity ywgWet gas specific gravity Turbulence factorVertical Depth (ft) AChange in a given parameter MMoody friction factor gGravity acceleration (m/s2) bReservoir thickness (ft) kPermeability (mD) kgGas relative permeability (mD) K]Friction losses KsGravity Forces IWell length (ft) HoOil molecular weight p Viscosity (cp) pDimensionless pressure pInitial pressure (psi) pReservoir pressure (psi) p]Reference pressure (psi) pw]Bottom-hole pressure (psi) pwhWell head pressure (psi) qFlow rate qgGas flow rate (Mscf/d)rcReservoir radius (ft) rwWell radius (ft) Porosity RgGas Condensate Ratio (scf/bbl) pDensity (kg/m3) sSkin tTime (h) tDimensionless time ITemperature (oF)zCompressibility (dimensionless) ReferencesAli, A.M., Falcone, G., Hewitt, G.F., Bozorgzadeh, M., Gringarten, A.C., 2005: Experimental Investigation of Wellbore Phase Redistribu- tion Effects on Pressure-Transient Data, paper SPE 96587 presented at the 2005 SPE Annual Technical Conference and Exhibition held in Dallas, Texas 9-12 October 2005. Cragoe, C.S., 1929: Thermodynamic Properties of Petroleum Products, Bureau of Standards, US Dep. of Commerce, 1929,Miscellaneous Publication No. 97, 22. Cullender,M.H.,Smith,R.V.,1956:PracticalSolutionsofGas-FlowEquationforWellsandPipelineswithLargeTemperatureGradients, Trans., AIME, 207 (1956), 281-87. D. DallOlio and L. Vignati, 1998: Well Heat Test Analysis: Save and Safe, paper SPE 39971 presented at the SPE Gas Technology Symposium held in Calgary, Alberta, Canada 15-18 March 1998. Fair, C., 2001: Is it a Wellbore of a Reservoir Effect?, Harts E&P. Fair, C., Cook, B., Brighton, T., Redman, M. and Newman, S., 2002: Gas/Condensate and Oil Well Testing-From the Surface, paper SPE 77701 presented at the SPE Annual Technical Conference and Exhibition held in San Antonio, Texas 29 September-2 October 2002. Govier, G.W., Fogarasi, M., 1975: Pressure Drop in Wells Producing Gas and Condensate, J . Cdn. Pet. Tech. (October-December 1975) 14, No. 44, 28. Hasan, A.R., Kabir, C.S. and Lin, D., 2005: Analytical Wellbore-Temperature Model for Transient Gas-Well Testing, SPE Reservoir Evaluation & Engineering, J une 2005. Nurafza, P.R., and Fernagu, J., 2009: Estimation of Static Bottom Hole Pressure from Well-Head Shut-in Pressure for a SupercriticalFluid in a Depleted HP/HT Reservoir, SPE 124578, paper presented at the SPE Offshore Europe Oil & Gas Conference & Exhibition held in Aberdeen, UK, 8-11 September. Peffer, J .W., Miller, M.A. and Hill, A.D., 1988: An Improved Method for Calculating Bottomhole Pressures in Flowing Gas Wells with Liquid Present, SPE Production Engineering 1988.Rzasa, M.J ., Katz, D.L., 1945: Calculation of Static Pressure Gradients in Gas Wells, Trans., AIME, (1945) 160, 100-05. Smith, R.V., 1950: Determining Friction Factors for Measuring Productivity of Gas Wells, Trans. AIME (1950) 73. http://www.spidr.com/oil-and-gas/Understanding-Wellbore-Cooling/subpage129.html