Gas Buildup in Pseudo-Pressure


Description

What permeability, mechanical skin and rate-dependent skin does a gas well show in two buildups after two rates? You paste each buildup's shut-in times and pressures and set the rates, the flow time, the gas and the rock. The sheet reads each buildup in the real-gas pseudo-pressure m(p) against pseudo-time, which make the gas flow equation behave like a liquid's, and fits its Horner line for k and the apparent skin S′. You read off k, the mechanical skin S and the rate-dependent coefficient D.

Parameters:

  • Reservoir, gas and well: pi, T, the gas gravity, h, φ, Swi, the rock and water compressibility, rw and the base pressure of m(p) (GPB_Pi, GPB_T, GPB_SG, GPB_h, GPB_phi, GPB_Swi, GPB_cf, GPB_rw, GPB_Pbase)
  • The two buildups: the two rates, the flow time before each, and the flowing pressure at each shut-in (GPB_q1, GPB_q2, GPB_tp, GPB_Pwf1, GPB_Pwf2; 0 takes the model's)
  • Analysis settings: the Bourdet window (GPB_L, 0.2), the radial-flow window in pseudo-time (GPB_From, GPB_To, 2.37 to 23.7 hr), how flat the derivative must be over it (GPB_Flat, ±10 %), and the round-trip tolerance on k and D (GPB_Tol, ±5 %)
  • Synthetic model: k, S, D and the wellbore storage coefficient the buildups are made from (GPB_ModelK, GPB_ModelS, GPB_ModelD, GPB_C)
  • Shut-in data: Δt and pws of each buildup, the regions GPB_Data1 and GPB_Data2

How to use it: Paste each buildup's Δt and pws over columns B and C. Type the flowing pressure at each shut-in into the two pwf inputs, set the rates and the flow time, and move the window onto the flat stretch of the derivative. Column A, the model's pseudo-time grid, is then unused, and the round-trip check doesn't apply. Two buildups after two different rates are needed for D. With only one, S′ is the result, and S = S′ − D·q needs a D from a correlation, such as the add-in's non-Darcy coefficient for a gas well. The sheet ships two synthetic buildups made from a model, so every result can be checked against the model that made it. A log-log plot draws the second buildup's Δm(p) and its derivative, a Horner plot draws both buildups with their lines, and a third plot draws S′ against rate.

Method: A gas well's skin grows with rate, S′ = S + D·q, because the flow near the well is turbulent; the two buildups' S′ therefore split into the mechanical skin S and the rate-dependent coefficient D. At pi the sheet takes Z by Dranchuk and Abou-Kassem at Sutton's pseudo-critical properties, and μg by Lee, Gonzalez and Eakin. It takes ct = (1 − Swi)·cg + cf with cg by its definition, and m(pi) from the pseudo-pressure integral (50 trapezoid steps). For each buildup row:

  • μg·ct at pws, from the same functions;
  • pseudo-time, Δta = Σ (μi·cti)·Δt/(μg·ct), the trapezoid of 1/(μg·ct) between rows, normalized at pi, from the flowing pressure at shut-in;
  • m(pws) − m(pwf), the Horner ratio (tp + Δta)/Δta, Agarwal's te = tp·Δta/(tp + Δta), and the Bourdet derivative of Δm against ln te.

The Horner line is the least-squares line of m(pws) against log₁₀ of the ratio over the window's rows. Its slope's magnitude is m, and its value at Δta = 1 hr is m1hr. Then k = 1637·q·T/(m·h), and S′ = 1.151·[(m1hr − m(pwf))/m − log₁₀(k/(φ·μi·cti·rw²)) + 3.23], the liquid Horner skin with m(p) for p and μi·cti for μ·ct. D = (S′2 − S′1)/(q2 − q1), S = S′1 − D·q1, and k is the two buildups' mean.

  • The synthetic buildups are the liquid solution written in m(p) and pseudo-time, which is what the two variables are for. The vertical-well model with storage and skin gives the buildup of a well with storage C and skin S + D·q, at q from 0 to tp and then shut in, on a grid of pseudo-times, with B = 1 and μi·cti in the diffusivity. Times 1422·T/(141.2·μi), that is Δm. A natural cubic spline through an m(p) table from 0.3·pi to 1.2·pi turns m(pws) back into pws. The real time Δt then follows by running the pseudo-time sum backwards along the model's pws. The analysis starts again from Δt and pws alone, as it would on a gauge's data.

Example: The gas and the well are the gas-inflow family's reference well: a 0.7 gravity sweet gas at 3,000 psia and 160 °F, 8 mD over 40 ft, rw 0.354 ft, a skin of 2, and D = 4.52 × 10⁻⁴ d/Mscf from the non-Darcy coefficient correlation with half the pay open. The sheet takes φ = 15 %, Swi = 25 %, cf = 4 × 10⁻⁶ 1/psi and C = 0.01 bbl/psi. At pi, Z = 0.8512, μi = 0.02131 cp, cti = 2.234 × 10⁻⁴ 1/psi and m(pi) = 6.135 × 10⁸ psi²/cp. The well flows 24 hr at 3,000 Mscf/d, and again at 8,000 Mscf/d, each to 2,737 and 2,107 psia, and each buildup is read for 23.7 hr of pseudo-time. Over the window from 2.37 to 23.7 hr, both derivatives are flat within ±1 %. Buildup 1 gives m = 9.591 × 10⁶ psi²/cp/cycle, k = 7.93 mD and S′ = 3.26. Buildup 2 gives m = 2.563 × 10⁷, k = 7.92 mD and S′ = 5.47. So k = 7.92 mD, D = 4.43 × 10⁻⁴ d/Mscf and S = 1.93. That is 0.9 % below the model's k, 2.0 % below its D and 0.07 below its skin, within the ±5 % tolerance. The small shortfall is the Horner line's, fitted over a buildup as long as the flow before it. Turbulence makes 65 % of S′ at 8,000 Mscf/d. Pseudo-time runs 0.6 % and 1.6 % behind real time at the end of the two buildups, more after the larger drawdown.

The sheet was checked against a standalone script. It rebuilds both buildups from the Engineering library's liquid PTA model and its Gauss-Legendre m(p), over the library's DAK, LGE and Sutton correlations, with m(p) inverted by bisection rather than a spline and cg by differencing DAK's Z. It then repeats the analysis with the library's Horner skin. It gives pwf 2,737.37 and 2,107.02 psia, slopes 9,590,592 and 25,629,117, k 7.93277 and 7.91599 mD, S′ 3.25768 and 5.47348, D 4.43159 × 10⁻⁴ and S 1.92821, all within 0.02 % of the sheet. Z and μg at pi are the gas-inflow sheets' (0.8512, 0.02131 cp), and the library's m(pi) is 613,556,320 psi²/cp, 0.007 % above the sheet's 50-step trapezoid. No printed two-rate gas buildup with pseudo-time and every input given was available, so the sheet is not compared with a published one.

Functions used

PO.PVT.Ppc.BySutton(SG_gas)

Calculates pseudo-critical pressure of hydrocarbon gas using Sutton (1985) correlation, [psi].

PO.PVT.Tpc.BySutton(SG_gas)

Calculates pseudo-critical temperature of hydrocarbon gas using Sutton (1985) correlation, [degR].

PO.PVT.Z.ByDAK(Ppr, Tpr)

Calculates gas compressibility factor (Z) using Dranchuk and Abou-Kassem (1975) equation of state, [dimensionless].

PO.PVT.Ug.ByLGE(Z, SG_gas, P, T_degR)

Calculates gas viscosity using Lee, Gonzales and Eakin (1966) correlation, [cP].

PO.PVT.Cg.ByDefinition(P, T_degR, SG_gas)

Calculates gas compressibility, [1/psi].

PO.Spline.Cubic.Interp(x_values, y_values, t)

Calculates interpolated value using natural cubic spline interpolation.

PO.PTA.Pw.VW(time, prod_data, B, Ul, C, Rw, S, Ct, porosity, h, K)

Calculates wellbore pressure drop for vertical well in infinite homogeneous reservoir, [psi].

PO.PTA.Horner.S(P1hr, Pwf, m, K, porosity, Ul, Ct, Rw)

Calculates the skin factor from the semilog straight line and abs(m); for a drawdown pass Pi as P1hr and the line's 1-hour pressure as Pwf, [dimensionless].

PO.DCA.Diag.Bourdet.SemiLog(x_values, y_values, t, L)

Calculates the classical Bourdet derivative dy/d(ln x) at point t, the log-domain estimator used for semilog and log-log well-test analysis. Multiply by LN(10) for a slope per log cycle.

Names (93)

NameRefers toComment
PO.PVT.PseudoPLAMBDA =LAMBDA(p, t_r, sg_g, p_base, steps, IF((p<p_base)+(t_r<=0)+(sg_g<=0)+(steps<1)+(p_base/PO.PVT.Ppc.BySutton(sg_g)<0.01), NA(), LET(ppc, PO.PVT.Ppc.BySutton(sg_g), tpr, t_r/PO.PVT.Tpc.BySutton(sg_g), n, INT(steps), dp, (p-p_base)/n, f, MAP(p_base+dp*SEQUENCE(n+1, 1, 0), LAMBDA(x, LET(z, PO.PVT.Z.ByDAK(x/ppc, tpr), x/(PO.PVT.Ug.ByLGE(z, sg_g, x, t_r)*z)))), 2*dp*(SUM(f)-(INDEX(f, 1)+INDEX(f, n+1))/2)))) Real-gas pseudo-pressure m(p) = 2*integral of p/(ug*Z) dp from p_base to p (psi2/cp), trapezoid rule in steps; Z by DAK and ug by LGE at Sutton pseudo-criticals. p, p_base psia (p_base >= 7), t_r degR, sg_g air = 1.
GPB_Pi $B$4 Initial reservoir pressure, pi, psia
GPB_T $B$5 Reservoir temperature, T, °F
GPB_SG $B$6 Gas specific gravity (air = 1), sweet gas, air = 1
GPB_h $B$7 Net pay, h, ft
GPB_phi $B$8 Porosity, φ, %
GPB_Swi $B$9 Water saturation, Swi, %
GPB_cf $B$10 Rock and water compressibility, cf + Swi·cw, 1/psi
GPB_rw $B$11 Wellbore radius, rw, ft
GPB_Pbase $B$12 Base pressure of m(p), psia
GPB_q1 $B$14 Rate before buildup 1, q1, Mscf/d
GPB_q2 $B$15 Rate before buildup 2, q2, Mscf/d
GPB_tp $B$16 Flow before each buildup, tp, hr
GPB_Pwf1 $B$17 pwf at shut-in, test 1 (0: the model's), psia
GPB_Pwf2 $B$18 pwf at shut-in, test 2 (0: the model's), psia
GPB_L $B$20 Bourdet smoothing window, L, cycles
GPB_From $B$21 IARF window from pseudo-time Δta, hr
GPB_To $B$22 IARF window to pseudo-time Δta, hr
GPB_Flat $B$23 Derivative flat within ± over the window, %
GPB_Tol $B$24 Round trip: k and D recovered within ±, %
GPB_ModelK $B$26 Model permeability, mD
GPB_ModelS $B$27 Model mechanical skin, S
GPB_ModelD $B$28 Model rate-dependent skin coefficient, D, d/Mscf
GPB_C $B$29 Wellbore storage coefficient, C, bbl/psi
GPB_Inputs $B$4:$B$29 The reservoir, the gas and the well, the two rates and flow time, the shut-in pressures, the window and tolerances, and the synthetic model
GPB_TR $B$32 Reservoir temperature, absolute, °R
GPB_Ppc $B$33 Pseudo-critical pressure (Sutton), psia
GPB_Tpr $B$34 Pseudo-reduced temperature, T/Tpc (Sutton)
GPB_Zi $B$35 Z-factor at pi
GPB_Mui $B$36 Gas viscosity at pi, μi, cp
GPB_Cti $B$37 Total compressibility at pi, cti, 1/psi
GPB_McI $B$38 μi·cti, the pseudo-time's reference, cp/psi
GPB_mPi $B$39 Pseudo-pressure at pi, m(pi), psi²/cp
GPB_Scale $B$40 Model Δm per liquid-solution Δp, 1422·T/(141.2·μi)
GPB_S1 $B$42 Model apparent skin, test 1, S + D·q1
GPB_S2 $B$43 Model apparent skin, test 2, S + D·q2
GPB_PwfModel1 $B$44 Model flowing pressure at shut-in, test 1, psia
GPB_PwfModel2 $B$45 Model flowing pressure at shut-in, test 2, psia
GPB_PwfU1 $B$46 pwf at shut-in used, test 1, psia
GPB_PwfU2 $B$47 pwf at shut-in used, test 2, psia
GPB_mPwf1 $B$49 m(pwf), test 1, psi²/cp
GPB_McPwf1 $B$50 μg·ct at pwf, test 1, cp/psi
GPB_N1 $B$51 Points in the IARF window, test 1
GPB_Flat1 $B$52 Derivative spread over the window, test 1, %
GPB_M1 $B$53 Semilog slope in m(p), test 1, psi²/cp/cycle
GPB_m1hr1 $B$54 Line's m(p) at Δta = 1 hr, test 1, psi²/cp
GPB_k1 $B$55 Permeability, k = 1637·q·T/(m·h), test 1, mD
GPB_Sa1 $B$56 Apparent skin, S′ = S + D·q1
GPB_TaRatio1 $B$57 Pseudo-time over real time at the last point, test 1
GPB_mPwf2 $B$59 m(pwf), test 2, psi²/cp
GPB_McPwf2 $B$60 μg·ct at pwf, test 2, cp/psi
GPB_N2 $B$61 Points in the IARF window, test 2
GPB_Flat2 $B$62 Derivative spread over the window, test 2, %
GPB_M2 $B$63 Semilog slope in m(p), test 2, psi²/cp/cycle
GPB_m1hr2 $B$64 Line's m(p) at Δta = 1 hr, test 2, psi²/cp
GPB_k2 $B$65 Permeability, k = 1637·q·T/(m·h), test 2, mD
GPB_Sa2 $B$66 Apparent skin, S′ = S + D·q2
GPB_TaRatio2 $B$67 Pseudo-time over real time at the last point, test 2
GPB_k $B$69 Permeability, k (mean of the two), mD
GPB_kh $B$70 Permeability-thickness, kh, mD·ft
GPB_D $B$71 Rate-dependent skin coefficient, D = ΔS′/Δq, d/Mscf
GPB_S $B$72 Mechanical skin, S = S′1 − D·q1
GPB_TurbShare $B$73 Turbulence's share of S′ at q2, %
GPB_kErr $B$74 k against the model, %
GPB_SErr $B$75 S against the model
GPB_DErr $B$76 D against the model, %
GPB_RateCheck $B$77 Rate check verdict
GPB_RateCheckNote $C$77 Rate check comment
GPB_WindowCheck $B$78 IARF window check verdict
GPB_WindowCheckNote $C$78 IARF window check comment
GPB_TurbCheck $B$79 Turbulence check verdict
GPB_TurbCheckNote $C$79 Turbulence check comment
GPB_TripCheck $B$80 Round trip on the synthetic data verdict
GPB_TripCheckNote $C$80 Round trip on the synthetic data comment
GPB_Results $B$32:$B$80 The gas at pi, the model at shut-in, each buildup's Horner line, k and S′, the mean k, D and S, the round trip and the checks
GPB_R0 $B$85 Rankine at 0 °F, °R
GPB_Steps $B$86 Trapezoid steps of each m(p) integral
GPB_K1422 $B$87 Gas pressure group, m(p) form
GPB_K1412 $B$88 Liquid pressure group
GPB_K1637 $B$89 Gas semilog permeability constant
GPB_Mega $B$91 Pseudo-pressure in one 10⁶ psi²/cp, psi²/cp
GPB_Eps $B$90 Slack on the window's ends, for rounding
GPB_MpTable $D$4:$E$19 Pressures from 0.3·pi to 1.2·pi and their m(p), which the model's spline reads backwards
GPB_PTab $D$4:$D$19 The m(p) table's pressures, psia
GPB_MTab $E$4:$E$19 The m(p) table's pseudo-pressures, psi²/cp
GPB_Rates1 $A$96:$B$97 Test 1's model: q1 from 0 to tp, then shut in
GPB_Rates2 $D$96:$E$97 Test 2's model: q2 from 0 to tp, then shut in
GPB_Lines $A$102:$C$110 Each buildup's Horner line as Δm(p), from a ratio of 1 to the data's largest
GPB_Curve $A$114:$C$116 The line S + D·q from q = 0, and the two buildups' apparent skins on it
GPB_Data1 $A$122:$C$149 Buildup 1: the model's pseudo-time grid, and Δt and pws (the model's until you paste a gauge's)
GPB_Analysis1 $D$122:$K$149 Buildup 1 per row: μg·ct at pws, pseudo-time, m(pws), Δm, the Horner ratio, te, the Bourdet derivative and the IARF window
GPB_Data2 $A$154:$C$181 Buildup 2: the model's pseudo-time grid, and Δt and pws (the model's until you paste a gauge's)
GPB_Analysis2 $D$154:$K$181 Buildup 2 per row: μg·ct at pws, pseudo-time, m(pws), Δm, the Horner ratio, te, the Bourdet derivative and the IARF window

Details

Inserts181 rows × 11 columns
CategoryPressure Transient Analysis
For agents and tools Descriptor (.md)
Tags:
gaspseudo-pressurepseudo-timebuildupHornerrate-dependent-skinnon-Darcywell-testing

Need a custom calculation? Request a Blueprint