BABYL OPTIONS: Version: 5 Labels: Note: This is the header of an rmail file. Note: If you are seeing it in rmail, Note: it means the file has no messages in it.  1,, Return-Path: Received: by lamont.ldgo.columbia.edu (4.1/SMI-3.2) id AA00369; Tue, 7 Nov 95 10:53:54 EST Date: Tue, 7 Nov 1995 10:53:40 -0500 From: Steven Nelson Message-Id: <9511071553.AA02931@rhob> To: 4d@lamont.ldgo.columbia.edu Subject: 4D Meeting Notes #2 X-Sun-Charset: US-ASCII Sender: owner-4d@ldgo.columbia.edu Precedence: bulk Reply-To: Steven Nelson *** EOOH *** Return-Path: Date: Tue, 7 Nov 1995 10:53:40 -0500 From: Steven Nelson To: 4d@lamont.ldgo.columbia.edu Subject: 4D Meeting Notes #2 X-Sun-Charset: US-ASCII Sender: owner-4d@ldgo.columbia.edu Precedence: bulk Reply-To: Steven Nelson 4-D Project October 23, 1995 Report (Report #4) 1. 4-D Meeting - Penn State and Lamont Doherty met on October 23, 1995, to discuss the status of 4-D research. See Attachment #1, "Meeting Notes," for details. 2. Status of Data Acquisition - a. Shell, Lamont, and Penn State have signed a data license agreement for the South Timbalier 295 field. Paper and digital logs are being sent this week from Shell to Penn State for development of a database for Penn State and Lamont to work with. b. Shell's Grover and Timberwolf 3-D surveys are being sent in segy form to Columbia forloading into Landmark (.3dv) format. Lamont has loaded these surveys into AVS and is presently working on converting the coordinates to a Landmark Project. c. Western, Lamont, and Penn State are currently negotiating release of a 3-D seismic data set over the South Timbalier 295 field. The data license agreement will have the same wording as the Shell/Lamont/PSU data licensing agreement, and the data is expected at Lamont next week. d. Texaco has released to Penn State a seismic inversion of their EI 338 -339 3-D survey. e. We have received preliminary permission from Diamond/PGS for a copy of the 1994 reshoot of the study area for Lamont/PSU. There is a draft of a collaboration agreement that would have PGS joining the consortium. GECO so far has refused to participate if another service company is allowed in. We are reapproaching them from higher in the Schlumberger organization. 3. SEG Meeting A Success - We had the only 4D booth at the SEG in the Lamont booth (albeit way to the side in the university consortiums area), but were also live on the internet in both Netscape and Hot Java in the Sun and SGI booths as well. All indications are that it was a splendid show. Requests for participation information were received from Statoil, Hydro, Elf, Unocal , and Total. If any of these companies are interested, we will be discussing with the current participants the data gifts that they would be required to include in order to qualify. 4. AAPG Abstracts Submitted A total of 7 AAPG abstracts were submitted by the Penn State and Lamont groups. They are listed in Attachment #2. 5. Important Dates - November 27 - Next monthly meeting. East Mountain Inn, Wilkes-Barre, Pennsylvania (Peter Flemings will reserve and pay for meeting room). Industry welcome to attend. January 8 - Corporate 4-D meeting in New Orleans. Please plan to attend. Any volunteers for hosting the meeting among our New Orleans companies: Chevron, Shell, or Texaco? Attachment #1: Summary of Lamont-Penn State Meeting on 10/23/95 The second monthly Lamont-Penn State 4-D Seismic meeting was held on October 23, 1995 at the East Mountain Inn in Wilkes-Barre, Pennsylvania. Presentat the meeting were Roger Anderson, Wei He, and Albert Boulanger of Lamont Doherty, and Peter Flemings, Andrew Hoover, Tucker Burkhart, and Steve Nelson ofPenn State. The meeting began with brief reports of each groups progress. Penn Statehas picked tops and bottoms for the LF and JD sands, examined sand character and its relationship to structure, mapped the LF and the JD on 3 seismic surveys, further investigated amplitude maps, and developed their own `normalization' procedure - which is simpler than the Lamont technique - for the seismic surveys. Lamont Doherty has normalized and differenced the Pennzoil EI330 survey and the EISOI survey in fault block B and found that they can adequately correct forthe banding present in the EI330 survey. The majority of the technical discussion centered around methods for the`normalization' of seismic surveys, perhaps the most critical task for insuring a successful 4-D project. Tucker Burkhart and Peter Flemings have spent a great deal of time in investigating this problem. By approximating the distribution ofamplitudes as log-normal, the mean and standard deviations of different surveys can be scaled to be the same. This procedure insures that although the number ofoccurrences of a single amplitude may not be the same, that the relative numbersof amplitudes within the survey are the same. Wei He and Liqing reviewed the more complex cross-correlation and spectral matching procedures Lamont has developed. They discussed the Lamont procedures for the 4 problems involved in seismic normalization: power spectra, amplitude, location, and phase problems. The first step in normalizing seismic surveys is insuring that the power spectra are the same. By filtering out the high end frequencies using a different box filter in each of three "depth" zones(0-2, 2-4, and 4-6 seconds TWT), the power spectra for each survey can be normalized. The amplitudes are then shifted with a fractional multiplier to insure the same amplitude distributions. The locations of the seismic data must also be aligned. A rebinning of the data is done on each survey to align the bins in the North-South orientation of the EISOI survey. Following the technical discussion, the list of responsibilities and goals that are listed on the following page was agreed upon. Responsibilities for Penn State: 1. Transfer LF sand top and bottom picks in an OpenWorks format to Lamont Doherty 2. Transfer interpreted well porosity, lithology, and water saturation data to Lamont Doherty 3. Interpret water saturation and sand/shale ratio data at half-foot intervals in well logs and transfer this information to Lamont Doherty 4. Examine possible `delta-front' lobes of the JD sand and calculate net sand `by lobe' 5. Copy Hart and Alexander papers and paper copies of JD and LF sand top and bottom picks and send to Lamont Doherty 6. Examine saturation data with corresponding permeability and porosity data 7. P.B. Flemings contact Chevron and Pennzoil 8. P.B. Flemings contact Wei He about seismic/velocity problems 9. Map velocity (v), density (r), porosity (f), permeability (k), water saturation (Sw), volume of sandstone (Vs), volume of shale (Vsh) , and structure for the JD, KE, and LF sands. 10. Map the KE on all 3 seismic surveys 11. produce velocity plots for each seismic survey (velocity differencing) 12. Produce amplitude and reflection strength maps for the JD, KE, and LF 13. Get TDT logs for block 330 from Pennzoil 14. Get production data for block 338 (Paul Hicks) 15. Load South Timbalier data (well data) 16. Write monthly report 17. Ask Shell Offshore for South Timb. production, TDT, deviation, mud weight, and leakoff data Responsibilities for Lamont Doherty: 1. Transfer rebinned `normalized' and `un-normalized' EISOI and EI330 surveys in fault block B to Penn State in Landmark format 2. Rebin and normalize the EISOI, EI330, and Roice 8 surveys within the cross over area, then transfer to Penn State 3. Investigate the problem of banding parallel to shot lines of the EI330 seismic surveys 4. Acquire seismic data from Diamond 5. Inquire with Pennzoil about getting 32 (or 16) bit reprocessed Western Data for the EI330 6. Solve 32 bit Landmark problem with Shell/Pecten 7. Load EI330 to AVS; complete "big" (fault block south of the f-fault) and "little" (fault block south of the g-fault) LF differencing 8. Load/Deliver the South Timbalier seismic data to Penn State 9. Get contracts signed 10. Implement software mods with industry test sites at Shell BTCC and Exxon EPR 11. Create segy and 32 bit Landmark project for South Timbalier 12. As soon as ST seismic surveys are in hand, begin working toward acquiring near/far stacks in South Timbalier datasets Attachment #2: Abstracts submitted to AAPG The following seven abstracts, in alphabetical order by author, were submitted to AAPG for the annual meeting in San Diego, May 1996. Visualization of Hydrocarbon Drainage using 4-D Seismic Techniques ANDERSON, ROGER N., BOULANGER, ALBERT, HE, WEI. LIQING XU, Lamont-Doherty Earth Observatory, Columbia University, Palisades, NY, 10964 We have developed 4-D volume visualization algorithms to identify significant seismic amplitude interconnectivity and changes over time that result from active fluid migration. To visualize this 4-D system, we have developed AVS-based software that uses multiple 3-D seismic surveys done several years apart over the same blocks. Normalization, cross-correlation, and differencing produces images of drainage recorded in seismic attribute changes over time. We have applied these 4-D imaging techniques to the Eugene Island 330 Field of offshore Louisiana, central Gulf of Mexico. Three main producing reservoirs were examined, the JD, KE and particularly, the LF which is the most prolific oil producer in the field. Dim-outs were detected where production replaced oil with water during the interval of investigation (1985-1992), and seismic amplitude increases were observed where gas/oil ratios increased during production. These changes were then tested against time-dependent water saturation measurements made by pulsed-neutron logs through casing and production histories. Visualization of 4-D seismic monitoring utilizing repeated 3-D seismic surveys promises to identify missed hydrocarbon zones and provide enhanced production control in the future. 4-D Seismic Interpretation Technologies, and their Application to the Eugene Island 330 Field of Offshore Louisiana. ANDERSON, ROGER N., BOULANGER, ALBERT, HE, WEI. LIQING XU, Lamont-Doherty Earth Observatory, Columbia University, Palisades, NY, 10964 We have developed 3-D volume processing and attribute analysis algorithms to identify significant seismic amplitude interconnectivity and changes over time that result from active fluid migration. To accomplish this 4-D imaging, we use multiple 3-D seismic surveys done several years apart over the same blocks. We have applied these 4-D analysis techniques to known production from the most prolific Pleistocene oil field in the world, the Eugene Island 330 Field of offshore Louisiana, central Gulf of Mexico. Three main producing reservoirs were examined, at 4500', 5400' and 7200'. Dim-outs were detected where production depleted oil and gas during the interval of investigation, and amplitude increases were observed where gas/oil ratios increased during production. The "oil/water contact" movement was easily detected by the 4-D technique. When combined with active pressure and temperature monitoring, repeated 3-D seismic imaging of producing fields promises to identify missed hydrocarbon zones, and to provide the critical production management information of the future. 4D Seismic Analysis of the LF-sand, EI-330 Field, Offshore Louisiana BURKHART, TUCKER, Pennsylvania State University, University Park, PA; ANDREW HOOVER, Pennsylvania State University, University Park, PA; STEVEN E. NELSON, Pennsylvania State University, University Park, PA; and PETER B. FLEMINGS, Pennsylvania State University, University Park, PA We interpret differences in amplitude maps of the 'LF' sand (Eugene Island Block 330, offshore Gulf of Mexico) from 3 different seismic surveys to reflect changing acoustic behavior due to drainage with time. Amplitude histograms of the extracted surfaces are log-normally distributed with different means and standard deviations. These data were filtered to have equivalent log-normal distributions. The re-scaled amplitude maps were then differenced to examine changes in amplitude behavior through time. Most zones which are interpreted to have been water swept based on production data, show an amplitude decrease with time. However, several locations in this water swept zone, show high amplitudes with no decrease in amplitude strength with time. These zones could represent areas of low permeability strata which trap by-passed pay. Alternatively, these may represent bright spots due to static effects such as depositional fabric, or diagenesis of the reservoir and/or overlying seal. These preserved amplitude highs are particularly pervasive adjacent to a major antithetic fault bounding the region. We are currently constructing maps of saturation, density and velocity to further constrain the temporal signals observed in the seismic reflection data. Quantifying One and Two Dimensional Lateral Heterogeneities in Fluvio-Deltaic Reservoirs using 3-D Seismic Data DESHPANDE, ANIL, Pennsylvania State University, University Park, PA; PETER B. FLEMINGS, Pennsylvania State University, University Park, PA; and JIE HUANG, Exxon Production Research Co., Houston, TX We document scale-invariant statistics and strong anisotropy in rock properties from well log and 3-D seismic data in fluvial/deltaic reservoirs in the EI 330 Field, Gulf of Mexico. The scarcity of well log data in the lateral direction necessitates the use of lower resolution seismic data to quantify lateral heterogeneity. Spectral analysis of two dimensional seismic horizon slices and one dimensional traces extracted from the 3-D data cube reveal scale invariant behavior with a characteristic correlation parameter (b) in both dimensions. This parameter captures the degree of correlation in profiles or surfaces (b=0 for white noise). Two dimensional analysis of the horizon slice indicates a (b) value of approximately 2.0 while analysis of one dimensional profiles from the same slice reveal an anisotropy along depositional strike and dip with (b) values of 1.6 and 2.1, respectively. The higher correlation observed in the direction of stratigraphic dip may reflect the stratigraphic fabric associated with channel systems. This one dimensional lateral variability in seismic data also matches the variability in higher resolution horizontal well log data suggesting a scale invariant behavior over approximately 3 orders of magnitude (1-1000 feet). Different depositional environments within specific systems tracts may have characteristic correlation parameters that provide insight into one and two dimensional lateral variations of reservoir heterogeneity. Correlation parameters obtained in these environments may then be used to quantify rock fabric and provide constraints in the simulation of rock property fields in the inter-well region. State of Stress in a Plio-Pleistocene Gulf Coast Growth Fault: Implications For Fracture Driven Fluid Flow FLEMINGS, PETER B., Pennsylvania State University, University Park, PA; MARK, D. ZOBACK, Stanford University, Stanford, CA; and ROGER N. ANDERSON, Lamont-Doherty Earth Observatory, Palisades, N.Y. Buoyant forces from trapped hydrocarbons are equal to the least principal stresses in the EI-330 field (offshore Louisiana, Gulf of Mexico); we interpret that this zone is in dynamic equilibrium where hydrocarbons are contained by the stress field and where migration occurs by hydraulic fracturing of this dynamic seal. The Pathfinder well was drilled in 1993 to measure in-situ properties of a growth fault inferred to be a major hydrocarbon migration pathway. Fluid pressures rise from hydrostatic at the surface to 85% of lithostatic at 8000 feet. The minimum horizontal effective stress (Shmin-P), as measured in a series of hydrofracture experiments, first rises and then decreases with depth in the borehole. In the vicinity of the fault (8000 feet) Shmin-P is ~375 psi in the fault zone and ~575 psi in shales immediately above the fault zone. The fault zone is thin (1 -10 cm) and has micro-darcy permeability although several thousand feet of displacement occurred across it. Large vertical fractures are present immediately above the fault zone itself. 1800 foot hydrocarbon columns are present, abutted against the growth fault, in the vicinity of the stress measurements. The buoyant forces from these trapped hydrocarbons are approximately equal to the least principal effective stresses measured along the growth fault. The vertical fractures present in the hanging wall are suggestive of vertical hydrofracturing. These results imply the EI-330 growth fault zone is a dynamic system where fluid flow is dominated by fracture permeability as the pore pressure approaches the least principal stresses present. The Evolution of Pressure and Stress in a Plio-Pleistocene Growth-Fault Eugene Island Block 330, Offshore Louisiana GORDON, DAVID S., Pennsylvania State University, University Park, PA; and PETER B. FLEMINGS, Pennsylvania State University, University Park PA Finite-element modeling of compaction driven fluid flow in a deforming fault system is used to examine the hydrodynamic behavior of growth faults. The model is constrained by observations of the EI 330 field (Offshore Louisiana). Strata are comprised of a regressive sequence which grades from hard geopressured prodelta sediments to a transition zone of shelf margin deltaic sediments to hydrostatically pressured fluvial sediments. There are sharp horizontal pressure gradients across the growth fault. We model the observed stratigraphic evolution, and we assume permeability to be a function of lithology and compaction driven decreases in porosity. Low permeability sediments correspond to shale dominated prodelta facies, and higher permeability sediments correspond to delta front and fluvial facies. Model results show that high porosities and overpressures are maintained throughout the evolution of this basin in the pro-delta strata due to the inability of the low permeability sediments to expel formation fluids. The fluid pressures in the model vary from hydrostatic at the surface to lithostatic in the prodelta sediments. In accordance with the observed pressure field, the model shows large lateral pressure gradients across the growth fault. A key component we are now incorporating is fracture permeability where the permeability is proportional to the ratio of effective stresses (as constrained by stress measurements). We are now using this permeability relationship to address the role of the faults in fluid migration. These models provide insight into the evolution of geopressures and the role of hydrodynamic processes in the migration of hydrocarbons. 3-D Finite Element Seismic Modeling of Hydrocarbon Drainage in a Gulf Coast Mini-Basin: The Role of Seismic Modeling in 4-D Seismic Technologies HE, WEI, ANDERSON, ROGER N., and WANG, XUEFEN, Lamont-Doherty Earth Observatory of Columbia University, Palisades, NY; and TENG, YU-CHIUNG, Columbia University, New York, NY 4-D seismic technologies are to promote hydrocarbon recovery efficiency. Carefully derived seismic amplitude differences from two time-elapse 3-D seismic surveys are very useful in finding gas caps, and more importantly, bypassed hydrocarbons in existing oil fields. However, it is extremely difficult to quantify the newly discovered reserves. Using stochastic simulation in reservoir characterization to associate acoustic impedance inverted from 3-D seismic data to reservoir physical properties from well data, we are able to construct individual reservoir models in the Eugene Island Block 330 Field in the offshore Gulf of Mexico. These quantified reservoir models in terms of lithology, porosity, pore fluid pressure and fluid saturation are then used to form acoustic models by using modified Gassman's equations to accommodate multi-phase conditions. By applying 3-D finite element seismic modeling technique to each individual reservoir, the observed seismic amplitude differences are modeled accordingly. During the seismic modeling process, the production data, i.e., gas and oil ratio, pressure depletion, and oil-water contact are used to constrain our modeled differences. The preliminary results applied to two 3-D seismic sub-volumes observed seven years apart in the study area suggest that the delay in travel time caused by newly developed gas caps may substantially affect the seismic response of the reservoirs below them. Reservoir parameters deciphered from seismic modeling to fit the observed seismic differences must be subject to constraints from production history data, or the estimated reservoir parameters may not be realistic because the acoustic models are substantially controlled by many reservoir physical parameters.