Upload
others
View
4
Download
0
Embed Size (px)
Citation preview
Well Log Analysis & Consulting
About Us
Everett Petrophysics specializes in mineral-based well log interpretation For 28 years we have been providing consulting, coaching and analytical
services globally; previous 27 years spent with Schlumberger. Software has been developed to use Nuclear Spectroscopy (elements) and
Nuclear Magnetic Resonance (total and free porosity) to derive grain density, exponents m & n, Rw, permeability, porosity and saturation.
ROBERT (BOB) EVERETT P.Eng Over 55 years of experience using petrophysics analytical techniques to interpret oil and gas well logs Provided consulting services for Unocal, Z and S, Dresser, Baker Hughes, Schlumberger and the University
of Texas at Austin, as well as many international Energy companies B.Sc. In Mechanical Engineering
James (Jamie) Everett, B.A., M.Sc. 30 years experience developing highly specialized software Development of software for companies such as Verity, Autonomy and Hewlett Packard B.A. with High Honours in Physics as well as an M.Sc. in Biomedical Engineering
Analytical Services
Calculation of full mineralogy: full mineralogy is possible with elements from Nuclear Spectroscopy Important as the attributes of the minerals are used to provide a unique
solution and has value in obtaining the best petrophysical analysis possible accurate porosity from accurate grain density Accurate Sw over full range of Sw Better permeability without special shift-constants Provides results in same units as core analysis such as weight fractions of
minerals A system that can be relied on
Analytical Services
Calculation of full mineralogy: the best way to analyze logs
Aspect of analysis 2: rather than using Vsh, which is not measurable in routine core analysis, use Si, Ca, Al, Fe, Ti, S, Gd which are measurable. Convertible to minerals, CEC, m, n, Grain density etc.
Aspect of analysis 3: logs respond to minerals and elements; makes sense to use them
Our PhilosophyMinerals Are the Cornerstone Of
Efficient Log Interpretation
Conventional Approach vs Our Approach
Data Collected Guess Vsh,matrix Use Vsh Guess Sw
Data Collected Derive grain density
Derive m, n, Rw
Sw = core
Type of results: Vsh is empirical, wrong Sw conclusions, uncertainty is high
Type of results: Derive from core-based data, best conclusions, lower uncertainty
Conventional approach
Our approach
Why use guesswork from 1950’s when better methods available?
Example Case 1: Duvernay
Project Details: Duvernay Unconventional
Conventional Vsh results were valid only at low Sw
What services we provided: Offset Nuclear Spectroscopy added
What info we added: cation exchange capacity from each clay family results in
full range of Sw
CONCLUSION: Mineral-based interpretation works
Example Case 1: Duvernay
Conventional Vsh results were valid only at low Sw
Example Case 1: Duvernay Cation exchange capacity from each clay family results in full range of Sw
Example Case 2: Tight gas
Project Details: tight gas environment
Conventional Vsh results missed the zone
Mineral-based results located the zone
We added Nuclear Spectroscopy from offset well
CONCLUSION: Mineral-based interpretation works
Example Case 2: Tight Gas Conventional Vsh results missed the zone: Sw 100%, PHIE < 5%
Example Case 2: Tight Gas Conventional Vsh results missed the zone: Sw 100%, PHIE < 5%
NOTE: THE POINT IS SERIOUS ERROR IS POSSIBLE WITH A VSH MODEL. Depends too heavily on analyst’s preconceived ideas of pay/no pay.
WEAKNESS is Rw, m, n, Vsh, PermAll “guessed” with empirical models
Example Case 2: Tight gas
Mineral-based results located the zone: SW 30%, PHIT 12%: Productive with Frac.
Mineral-basedCalculation ok
elements/mineralsProvide cec, GD, perm& constraints for m, n.
STRENGTH is Rw, m, n, GD & Perm:All are Internally calculated,originally from core data
APPENDIX: Example How to obtain m, nSw calculation:
• The saturation equation used is called a Dual Water Equation, from the paper by Chris Clavier et al, (Ref 5).
• The components of the equation are:
m_zero, which is the cementation factor, dependent on m* (m_star), the Waxman-Smits cementation factor:
IF((m_star<=2.0356),(m_star/(0.1256*m_star+0.7781)),((m_star/(0.3764*m_star+0.2694)))), where
m_star=(1.653+(0.0818*(Surface area*RHOG)^0.5)), where
Surface area (SO) =sum of specific SO of each mineral. This is a very critical part of the calculation.
n_zero = tortuosity factor = m_zero =
IF((m_star <=2.0356),( m_star /(0.1256* m_star +0.7781)),(( m_star /(0.3764* m_star +0.2694))))
N_zero and m_zero are used in the Dual Water saturation equation; M_star would be used in the Waxman-Smits-Thomas equation, but we used the DW equation. In the lab, both DW and WS give the same results. However, DW also gives Swb, the clay water saturation. We use Swb as a quality control check because Swt cannot be lower than Swb.
APPENDIX: Example How to obtain SwWater Saturation Detail
Water saturation is calculated from the Dual Water equation, using cation exchange capacity to calculate Qv.
[See Clavier, C., G. Coates, and J. Dumanoir, 1977, the theoretical and experimental bases for the ''dual water'' model for the ... SPE Paper 6859, Society of Petroleum Engineers 52nd Annual Fall Meeting, Denver, Colorado, October 9–12, 1977. Dewan, J. T., 1983, Essentials of modern open-hole log interpretation: Tulsa, Oklahoma, PennWell Publishing Company, 361 p. ... Hill, H. J., O. J. Shirley, and G. E. Klein, 1979, Bound water in shaly sands: Its relation to Qv and other formation ...]
From the Petrophysics Designed to Honour Core (PDHC) program:
CEC = Sum(Wi*CECi)
Qv = (CEC/100*Rhog*(Rhob-1)/(Rhog-1)/Tpor)
M = salinity/Rhof_bw/58.45, where salinity is the formation water salinity and Rhof_bw is density of bound water, usually 1. [Molarity.]
W = (0.22+(0.084/M^0.5))
Swb = W*Qv
B = 0.03772*Temp_degF-0.6516
Cb = B*Qv/Swb
F = F_Ghanbarian or 1/TPORm_zero
Cw = 1/Rw_SP_used
TC_DW = ((Cb*Swb-Cw*Swb)/F)) [Term c in quadratic]
Co = (Cw/F)+TC_DW
Ro = 1/Co
Ct = 1/Rt
Swt = (Ct/Co)n_zero
Hence,
Swt = (Ct / ((Cw/F)+ ((0.03772*Temp_degF-0.6516*Qv/ (0.22+(0.084/ salinity/1/58.45^0.5))* (CEC/100*Rhog*(Rhob-1)/(Rhog-1)/Tpor) * (0.22+(0.084/ salinity/1/58.45^0.5))* (CEC/100*Rhog*(Rhob-1)/(Rhog-1)/Tpor) -1/Rw_SP_used * (0.22+(0.084/ salinity/Rhof_bw/58.45^0.5))
*Qv)/F)))) n_zero
APPENDIX: Example How to obtain Rw: step 1
The Rw_SP_USED is calculated in two parts. First an estimate is made, called RW_SP, from an estimated field value and a baselined SP. Second, this value is refined so that after a full ECS calculation is made, the CEC-Corrected Ro matches the recorded deep resistivity. The Rw process is repeated until a satisfactory match of Ro and Rt is obtained. One has to be careful as the density tool has a different vertical resolution than the deep resistivity. Hence, some experience is useful.
APPENDIX: Example How to obtain Rw: step 1
• Rw_SP is derived from an estimated field value and a based lined SP. The trick is how to baseline the SP.
• Other methods baseline the SP by drawing a line from whatever is considered shale, called a shale baseline.
• This method ‘calculates’ and average zero from the SP equation. This step is important as shales have an SP as they are never zero.
APPENDIX: Example How to obtain Rw: step 1a
• First, calculate temperature from this equation or any other that provides a geothermal reservoir temperature. Depending on circulation time, the geothermal temperature sis expected to be 10F to 20F above maximum recorded logged temperature.
• 0.0198*depth in ft + 42.8 degF • Modify as you see fit.
APPENDIX: Example How to obtain Rw: step 1b, SP_ZERO
• Next calculate the estimated field Rw using the temperature gradient; standard formula
• Rw_known = 0.025*(75+6.77)/(TEMP_DEGF+6.77)• Now calculate Rmf using the same formula
0.04*(87+6.77)/(TEMP_DEGF+6.77)• Then SP_ZERO = (61+0.133*TEMP_DEGF) *LOG(RMF/RW_KNOWN)+add• The “add” is a constant to make the average value zero. This SP_ZERO is
the key step.
APPENDIX: Example How to obtain Rw: step 1c, SP_shift & SP_baselined
• Next, calculate SP_SHIFT = SP + add2• Where add2 is zero to start with but will be changed later.• Now SP_BASELINED = SP_SHIFT-SP_ZERO• Finally RW_SP = RMF/(10^(SP_BASELINED/(-1*(61+0.133*TEMP_DEGF))))• Now test to see if Rw_SP comes close to Rw_KNOWN.• If not, supply a new add2 & iterate until they are close. See plot next.
APPENDIX: Example How to obtain Rw: step 1c, shift required
APPENDIX: Example How to obtain Rw: step 1c, shift made
Add2=+40
Now RW_SP_CALCset to RW_SP
APPENDIX: Example How to obtain Rw: step 2, compare Ro to Rt; re-shift
Dashed Rw_SP_USEDAfter “add2”changedso that Ro=~Rtat arrowi.e. at low Resistivity, water or shale,since Ro has CECcorrection.
APPENDIX: Example How to obtain Rw: step 2, compare Ro to Rt; re-shift
Dashed “Rw_SP_USED” after “add2”changedso that Ro=~Rt at blue arrow.i.e. at low Resistivity, water or shale,
since Ro has CEC correction.Now we have a Rw that can be used for theentire well as it is modulated by the SP.This is an oil-base mud & SP was predicted!
Conclusion
Why this information is important: log interpretation used to be more art than science;
times have changed.
Additional point: with Nuclear Spectroscopy and Nuclear Magnetic Resonance, better
science in interpretation has resulted: all parameters initially internally computed
without analyst bias.
When science-based interpretation is available, why not use it?The conventional Vsh approach was all we had before 1980s but now we have a better, more accurate way to provide Petrophysics Designed to Honour Core
Contact Information
Robert V Everett PetrophysicsConsulting, Teaching, Coaching
AddressPO Box 271 – 1589 Nurmi RoadMerville BCV0R 2M0
Robert: 250-442-9696 / [email protected]: 403-620-2403 / [email protected]