Field of the Invention

This invention relates to oil and gas exploration and production.

The invention is concerned with the provision of means to convert prepared seismic data to geology, via the normalised depth method, by quantifying and adjusting for differences between local and 'normal' geology, as the mechanism structure plus the minimum several components within to allow performance .

The invention also relates to the provision of means to work prepared seismic data and normalised depth geologic models to generate normalised depth petrophysical models, via modelling to quantify and infill corrected differences from a generic model data set.

Further, the resultant trace sample material is rendered cellular, to allow spatial definition of the base of sealing rocks, so that gross rock volumes of all structural and strat traps are quantified per project area. Further, trace samples of geologic output are rendered cellular, for source rock quantification and forecast of oil/ gas type generation and volumes .

BACKGROUND TO THE INVENTION

Concerning Exploration and Production of oil fields over 30 million barrels (mb) , E&P's big problem is that geoscience asset team interpretation of lithology & poro-perms (holes in rocks & connectivity) are risky, wasting on average some 50% of exploration & 20% of production cost. Figure G5 summarises what seems retrospectively obvious about parts 1-6 of new methods invented to double knowledge of poro-perms from seismic.

Figure G5 summarises what seems retrospectively obvious about the benefits of the present invention to double knowledge of poro-perms from seismic data. This is required because now, generation of poro-perm information is still mostly reliant on well data, interpolated to rather than generated from seismic data.

That oil fields invariably produce different quantities and over different times than was forecast before production, suggests that the initial geo-poro-perm models were incompletely proper. In exploration, rarely is well data available to define behaviour of all lithologies at all depths .

Current oil & gas E&P geoscience asset teams do not have the methods or means to sensibly quantify rock porosity and permeability at the resolution of seismic (some 10m3) in projects averaging some 500km3, for two reasons. Firstly, good seismic attribute and shape material cannot be yet processed into good material quantifying poro-perms, because depositional lithofacies and burial changes cannot yet be sensibly quantified. (Today's prior art works using local well data, where the assumption is that the local well -based geological model is valid for local geo/ rock physics modelling) . Secondly, per project, even if means existed to do this in a few cells, the finite current HR & IT resources would still need new methods to do the work in 500M cells per project. Each team member does not have and urgently needs a common denominator, to transform material worked and to be worked, such that each knows what others are doing, and the collective whole properly and harmoniously manage all risks.

Clastic and carbonate continental shelf sediments, compact (lose pore space) in burial, at relatively predictable rates, in the absence of locally anomalous stress, strain, temperature, pressure or fluid systems. See Figure 1. In such circumstances, as depth increases, sediment interval velocity Vint plus its rate of compaction are commonly expressed as Velocity at origin Vo plus associated compaction rate K, of Vint change per particular sediment lithology type, with depth .

So, in such cases, each of say 20 key lithologies behave with depth, in a manner where local Vint is a function of Vo + K, so lithology is quantif able. This does not assist E&P efficiency much, because of potential for error in every place of different geology. See Figure 2. No rules and tools yet exist to quantify all such potential for locally different geology, or to use this difference to cross-correlate the ^• normal' material, with the local geological circumstances, elsewhere .

Figure 3 summarises poro-perm behaviour of key lithofacies, with potential for one lithology to have similar poro-perms in different depth ranges.

Figure 4 shows that, by knowing +/- a few %, velocity interval Vint and depth, plus the depth difference between that and a generic normal material compilation, once can convert samples into the material to better quantify traps, reserves, recovery and risk. To do this requires 6 classes of work to be performed to prepare the following:

1. ^Λ Normal' model

2. Seismic for normalised G&G processing.

3. Geology, from 1, 2

4. Petrophysics from 1 & 2

5. Trap volumes

6. Oil or gas fluid corrections & risk of overpressure.

The present invention is concerned with the conversion of seismic trace data into a pseudo well sonic (interval, Vint) velocity log, by actions causing each sample to represent the apportioned, defined sum of a geologic model of depositional lithology plus compaction normal for that lithology, plus a residual representing depth normalisation of localised difference between current depth and the depth needed to quantify local compaction and digenesis.

The present invention is also concerned with the conversion of seismic trace data with use of trace velocity log, and normalised depth geologic model, into normalised depth petrophysical models, focusing on sensible detailing of porosity and permeability.

Subsurface information is worked by multidisciplinary asset teams, each using one of a few types of very similar IT workstations, where each project is worked without ability to quantify all relevant causes or effects of rock properties, (i.e. part subjectively) limiting capacity of any expert to understand or QC the whole work.

Seismic data now worked in oil and gas exploration and production allows per unit space, wide ranges of geological models to be derived, mostly because current depth per unit volume has not been depth normalised to show maximum effective compaction.

Subsurface information is worked by multidisc plinary asset teams, each using one of a few types of very similar IT workstations, where each project is worked without ability to quantify all relevant causes or effects of rock properties, (i.e. part subjectively) limiting capacity of any expert to understand or QC the whole work. Seismic now is rarely worked at its trace sample resolution of say 2 milliseconds, into lithology, poro-perms etc. When this is done, it invariably requires use of well data from local similar (?) geology, and much iteration via inversion procedures, to then have 50% probability of being wrong in complex exploration projects.

A primary problem that has limited progress in this task area is the fact that at any elevation in sedimentary basins, between say + 2000 metres and - 6000 metres, per very similar rock deposit, or lithofacies type, most elastics (sands, shale, mudstones) and carbonates (chalk, limestone) commonly exhibit wide ranges of velocity, porosity and permeability. So, who can say what 'normal' behaviour is?

A good correlation is reported to exist between depth and velocity when material is compacted in a rock mechanics lab, or when it is compacted in much of the shallower areas of North Sea, Norwegian offshore or USA Gulf Coast. Here, lithofacies/ velocity/ depth/ poro-perm relationships are fairly robust, in the absence of cause to make this not so. Per lithofacies, all evidence suggests that permeability is a close function of porosity. So, what is needed to cut risk of error is a complete understanding of all such interrelationships in a particularly simple geological set of circumstances plus a set of rules and tools to enable quantif cation of any differences in geological circumstances per other project area, whereby such differences are particularly assembled to enable cross referencing between the norm and the local .

Procedures known as ^λ Becvem' and 'Spiral' go some way to address these background problems . Need 1 is to have a generic 'normal' geologic and petrophysical listing in depth, against which all other geological project data can be compared and information derived, instead of having to make from direct sensed data a different listing per different geological basin, and then have to construct specific methods to quantify any differences between that direct sensed model and all other remote sensed geophysical data.

SUMMARY OF INVENTION

To overcome the problems outlined above, the present invention uses filters, rules, libraries of physical evidence and other means to break down seismic (velocity, density) data into its 3 -classes of geological components, namely depositional lithology, normal lithology compaction and any local abnormal compaction.

Interval velocity, Vint, as defined in co-pending Application No. (Case 1/2) is treated as the sum of velocity at datum zero, Vo, plus compaction normal, Kn. Vint is modelled as Vo + Kn plus or minus any Vint variation by abnormal burial, Vint of Ka. Quantified local presence / sequence/ sample or cell of Vint variation by abnormal burial Ka, usually requires a rebalancing of Vo + Kn to continue to fit Vint.

Below normal burial depths at which shale becomes a viable seal, variations in Vint between the several most common lithologies tend to exceed the resolution at which Vint is now quantifiable from seismic trace sample data. When lithology and Vint are known per sample, porosity and permeability can be charted. These are the minimum number of parameters needed for sensible classification of geoscience related quantification of seals, carrier beds, thus traps and their gross and net rock volumes. The extent by which all inter-relationships make sense to expert system of HR QC is made a relatively more objective definition of probability of correct prognosis.

The method of the present invention processes /samples (say 2ms) to differentiate any relative burial depth difference between normal generic depth set of geo and petro parameters and the project cell volume. Once the Vint of the project cell is understood as the sum of Vo + Kn adjusted in depth for the effect of Vint of Ka, then the project cell's lithology and poro-perms can be quantified by adjusting the generic normal data set. Therefore, these capacities include methods and means to quantify the separate and net causes and effects responsible for burial and or digenetic changes and their relative difference to the generic data set.

The present invention also includes the provision of means to alter output of several prior art methods, to provide relative burial depth equivalency differences plus 2 new methods to adjust for burial where water thickness varies and where sediments are relatively under-compacted, by cause of angle of slope.

Figure All:l: summarises key parts of the present invention and Figure All:4: summarises where the extra work of the 6 key parts fits within prior art workflows. Overall, it adds up to around 50% more work done per project, by machine batch processing, objectively, rather than heavily HR driven, subjectively. Further, about l/3 ^rd of labour intensive, prior art work is done by the new art, by machine in batch processing.

The present invention includes the use of high resolution seismic trace data, already prepared as velocity Vint, and as a detailed, 3 -class model of geology, explaining present depth, lithology, normal compaction for that depth plus via the depth normalisation process, detail of present equivalent depth of burial.

The present invention also provides means for use in a specific depth normalised domain to convert such seismic and geological models into first porosity, then permeability, thus local capacity to act as seal, reservoir rock or carrier bed. Filters, rules, libraries of physical evidence and other means can be used to break down seismic geology into the minimum petrophysical components needed for sensible classification of probability of occurrence.

Porosity per lithology is charted for all depth ranges, in a normal geological setting. Normal is defined as one where filters find no evidence of abnormal causes of burial changes. Permeability is charted as a function of lithology and porosity. Before potential of a trapped reservoir rock can be assessed, need is to quantify the geometry of seals, so this is done by charting (Figs 3, 4) , assuming pore fill is water or brine.

The present invention therefore includes the taking of project seismic trace sample depth, Vint, lithology and relative depth difference to generic normal geo-petro data derived by the above methods and quantifies poro-perms and per sample, potential to perform as a seal, or reservoir or source rock. It then converts the 2D vertical seismic profile/ trace data into cells, for 2 purposes. Firstly it assesses source rock potential, fluid type and generation volumes using a method of listing lithology, normal & abnormal burial controls and digenesis. It then quantifies the whole project rock volume to find the areal extent of all vertical seals, searching laterally and vertically to quantify the gross rock volume of all stratigraphic and structural traps. It generates material quantifying migration paths to a 3D trap, &/or to the surface .

The present invention also includes the use of different construction material, to generate different output by different means. It uses seismic Vint data, scaled to well & generic model data, then converted to geology. These different materials are worked to materialise poro-perms and key source rock properties in a different, depth normalised domain. It then converts 2D seismic to be used like 3D seismic 3D cells, accessible in time or depth, to also quantify vertical closure and gross rock volume of all structural /strat traps in project area plus detail of source rock volumes, type, generation, maturity, etc. The routine quantification processes, per seismic sample of interval velocity and lithology with or without use of well data using depth normalised methods provides extra means to quantify lithology. The prior art necessarily uses local direct sensed information or subjectively chosen analogies. An advantage of being able to quantify Vint plus lithology per seismic sample, with or without well data, is that this allows machine conversion with little further loss of accuracy of porosity then permeability, which themselves quantify rock capacity to act as seal or reservoir, and thus traps and their gross and net volumes. A further advantage is automated provision of a detailed geologic model of lithology- facies and its normal compaction, plus compaction abnormalities, massively adding to asset team knowledge and their capacity to quantify risk.

The routine quantification, per seismic sample of interval velocity, lithology and poro-perms with or without use of well data using depth normalised methods provides extra means to quantify lithology and poro-perms. The prior art necessarily uses local direct sensed information or subjectively chosen analogies.

In exploration for prospects over 20mboe (million barrels oil equivalent) advantage should improve success rates 50%, from one third to one half. Today, some 80% of this dry hole rate is caused by lack of means to sensibly map poro-perms. In production, the advantage should amend design specification of production facilities and wells to optimise flow rates and recovery, worth 10% of such cost. Today, activity is handicapped by lack of means to sensibly map poro-perms.

The present invention uniquely provides means to quantify all probable strat or structural trap gross rock volumes, plus source rock volumes, fluid types, maturity, migration paths, loss, etc from seismic.

INFORMA ION IN DRAWINGS

Concerning conversion of seismic data in projects of any oil & gas prospective geology, into geology of deposition and burial changes, referenced to a normal depth model of geophysics, geology and petrophysical material via quantification of depth differences between a generic normal compilation and the local material, an example of the present invention will now be described by referring to the accompanying drawings, in which:

• Figure Gl, normal v/d lithology looks like a fan, separating lithologies .

• Figure G2, porosity / lithology is quantified in normal depth, from velocity/ normal depth relationships.

• Figure G3 , shows fan shaped behaviour separating lithologies on poro-perm x-plot .

Concerning conversion of seismic data in projects of any oil & gas prospective petrophysical property behaviour, into poro- perms, seal, reservoir, carrier bed properties, referenced to a normal depth model of geophysics, geology and petrophysical material via quantification of depth differences between a generic normal compilation and the local material, an example of the present invention is described by referring to the accompanying drawings :

Processing petrophysical properties from seismic by normalisation:

Assuming brine or water pore fill, knowing relative normalised depth differences, construct per trace sample,

• Porosity from lithology & Vint,

• Permeability from porosity & lithology

• Seal/ reservoir/ carrier bed properties from poro-perm x plot.

• Source potential from thermal, depth, lithology, +/-age data as available

DETAILED DESCRIPTION

A bridge is constructed spanning seismic geophysics & geology. Lithology i) , + compaction normal ii) have fixed normal relationships via Vo + K. In the generic set Vint velocity = Vo + K. In other sedimentary basins, therefore, Vint = (Vo + K) +/- abnormal burial changes iii) . So, any change in i) or ii) must change those relationships, and any change in iii) must also change either i-ii) , or Vint. Geologists must QC-QA split of Vint into a sensible i-iii) geo-model /cell, or agree with geophysicists a Vint / cell, allowing this.

Systems Quantify separate & net effects of all key causes of abnormal iii) burial- digenesis /lithology per cell, e.g. non-vertical stress, inversion, slope extension, water depth, fluids etc. They then quantify depth equivalent difference in burial / lithology between project cells & the generic normal data.

Primary deliverable for normalised G&G processing of seismic is lithology, per seismic trace sample. This is derived by having generated the difference in depth, if any, required to normalise the project work cell, in which time velocity and depth are known, to position it at the depth of burial in the normal G&G property situation, which corresponds to the project cell data.

There are 4 stages of methods and means required to construct the primary deliverable.

Work stage 1 processes lithology and compaction, assuming burial/ digenesis is as per normal. Normal here includes fluid brine fill.

Work stage 2 part 1 filters stage 1 seismic and G&G material for material separately providing evidence that model 1 G&G is locally wrong.

Work stage part 2 constructs evidence of net change per cell, of all separate causes, at sequence vertical resolution, within the velocity dept^h domain, and converts this into a depth adjustment, positive or negative, in fit relative to the generic norm.

Work stage part 3 converts project data into geology at sequence resolution, optionally linking with default systems to minimise risk that lateral and vertical variations in the geology are improbable.

Work stage 4 converts the above at trace sample resolution, with time averaging, to ensure that sequence thickness is proper, and that all sub sequence units are proper in respect of their time, velocity, depth, lithology Vo, Kn & vint of a.

The present invention allows seismic & seismic geology to be converted into poro-perms via knowledge of the difference in depth between local spatial model and the generic normal model. Each trace is pseudo composite log, displaying 2-4D seismic in time or depth, as AI, Vint, Vo, K normal, Vint of K abnormal, and as Depth difference to generic, porosity, permeability, seal & reservoir via X-plot. Default systems are optionally used to ensure multidisciplinary sense at and between every sample & cell.

Primary deliverable for normalised G&G processing of seismic is petrophysical poro-perms, plus associated quantification of seal, reservoir and carrier bed properties.

Figure All-2 summarises general E&P G&G workflow, leading to efficiency noted

Figure All- 4 summarises key parts of extra rules and tools to increase E&P efficiency

Reference should be made to the method described in copending Application No. (Case 1/2) which provides input material as a full set of generic material representing all geological depositional rocks at all basin depth, (digitally as Vo) in one common geologic set of burial changes (digitally as Kn) plus means to rescale this where local project well data is available, to define and fit local project burial circumstances.

The method described in said co-pending application provides input material as a 2-4 D seismic + processing velocities, which is worked to separate as far as sequence layer resolution allows, all significantly different rock homogeneities, and then refining accuracy of definition of their pattern, character, attributes, velocities Vint, and geometry in time and depth. It also also provides a first pass geologic quantification and QC of geology of deposition, as lithology as Vo and burial change as Kn, assuming it to be as per generic norm, optionally as locally (Vint of Ka) amended by well control. Figure Gl shows normal geology, with depth/ velocity separation by several % of lithologies, fanlike. Seismic resolution of velocity & depth is a few %. So, in such 'normal geology, seismic velocity/ density largely explains lithology & compaction.

The question thus answered by the present invention is 'can all other basin sediments with different burial conditions relative to the above, be worked to show relative depth of burial difference (delta D)?' If so, it is consequential that knowing the depth difference equivalency, one can back out lithology & poro-perms, from known velocity & dept .

There is no prior art common denominator between seismic/ geology & petrophysics . Knowledge of local seismic delta D relative to generic norm is a very low common denominator, LCD.

In the present invention, Vint from seismic, is further QC'd, by processes addressing the fact that Vint should be very accurate, and may be helped to be so where geology is materialised (LCD) as the sum of deposition Vo + Kn normal burial change + Vint of Ka abnormal burial changes, expressed as D.

So, Geophysicists , Geologists structural & stratigraphic , Petrophysicists & Managers can and should all bring and harmonise their separate multidisciplinary expertise to properly balance the separate parts of the total within the total that is almost always known with great accuracy. The biggest set of unknowns facing asset teams concern their capacity to quantify and QC that Delta D's behave sensibly with age, in how episodes of compression, inversion, extension effect contemporaneous and older sediments, but not younger sediments. Then, the local well data model can be calibrated to the generic normal model, to derive full information at all depths, all space, by integration with the seismic derived model .

The present invention thus includes the provision of a processor equipped for receipt of the various types of input material, to then process the key causative controls acting to compact sediments. This is done at seismic trace samples, or groups of, +/- cells. The primary end output is delta depth, relative to pre defined normal geology.

Fig 3:4: is a General Table in which the top left is the initial model worked in and input and the top right summarises the primary cause controls of property change in burial, and system of means to quantify. The lower middle is means to confirm that Vint +/- Vint of delta D has high probability as a valid Vo + K normal model.

Apparent Geology compound system.

Mapping of probable apparent geology in sequences provided, at resolution optionally in the horizontal ranges of about 100m2 to 25m2. The method in the co-pending application referred to above derives and records depth differences between generic normal defined apparent lithology and associated properties at particular depths, and similar properties per similar lithology in local project wells derived lithologies. Constrained are velocities (minima, maximums, gradients) to represent what is possible for the apparent lithologies and depths, recording adjustments made.

The present invention relates to the structure within which the industrial process of working out delta D, relative to the norm. This structure has entry and storage and work and exit areas. Next, it is the several sets of processes that quantify the local effect, separately and collectively, of the several different geologic processes that cause rock property changes in burial. Of these, enough are described herein in enough detail to allow experts in the art to make systems do useful industrial work. It is noted that whereas each and collectively, these components do the job at say >80% efficiency, those with better logic and more time have much opportunity to design systems to achieve better results.

The main structure ensures the sensible behaviour of Vint = Vo + Kn +/- Vint of Ka, where the latter is converted into a relative burial depth difference with generic norm, (&/ or generic norm scaled by locally available well data) . The method used thus models seismic to velocity Vint & depth, then computes lithology & compaction Vo + Kn, as if geology is as per generic norm, (or as scaled by locally available well data) , then this geological model is updated to account for material evidence of different compaction/ digenesis.

The depth difference is quantified from Vint of Ka, after Vint of Ka, abnormal compaction, is used to amend as appropriate Vo +Kn and thereby local cell lithology and generic model compaction.

In Figure 1-9, the sloped line represents 'normal' compaction of a particular lithology, and the intersection of the vertical and horizontal lines represents the different velocity, depth of the same lithology. So, one should look for evidence to explain this. Likely examples are in the form of structural inversion, to cause the depth difference, &/or non vertical stress to increase the local velocity. To find such corroborative evidence, needs particular · tools .

The need is for systems to quantify separate and net normal and abnormal burial alterations. The first part of the work flow of the present invention uses high resolution sequence material .

The primary work structure calls upon the several 'filters to quantify burial changes & anomalies, per sequence', surface down, then to store such material. This is done at the highest chosen horizontal resolution per sequence. Each component's material evidence of such changes (Vint of Ka, & associated delta depth) , see fig abnormal Vint & delta D, should then be processed to ensure that behaviour seen passes default settings of possible lateral and vertical behaviour .

Per hi-res sequence cell, then need exists to quantify which effects from the several filters act

• Concurrently, where any one cause has cut pore space, which a different cause having similar effect won't decrease pore space further.

• Consecutively, where effect of 2 or more change causative processes is largely summed. For example, water depth increases and sediment slope angle increases both may act to cut rate of compaction with depth.

• To cancel out one effect by another effect, e.g. where earlier lifting up by say 1000m (inversion) is made irrelevant by subsequent burial of say 2000m.

The set of operations to do this NENDA, works out net delta D per sequence cell, using rules, after the separate other filters/ algorithms have assembled their material. Derive evidence of Kn & Ka at sequence resolution using seismic interpretation data, and well data as available. These are mostly functions of normal compaction, non vertical stress compaction, structural short and long wave inversion, faulting, pressure and fluid and thermal anomalies. Separate and net effects variously quantified as below described then need to be expressed in terms of relative depth difference to the generic normal behaviour. This then requires tools to quantify Depositional systems. Litho-facies are a function of lithology, environment, structural setting and climate, all to be defined as far as possible per geo-cell. Derive evidence at sequence resolution using seismic interpretation data, geometry & seis strat, and well data as available. Commence by subtracting burial alterations from apparent geology then re- quantify the adjusted lithology that best fits Vint = Vo + Kn +/- any Vint of Ka. Then map lithology, depositional environment, and structural setting, and palaeoclimate data if available. Means allow QC that each set of properties are possible and pass defaults in 3D space & time.

A resolution error system is preferably provided to allow per sample or cell seismic Vint of generic normal geology Vo + Kn & abnormal Burial changes to equal resolution error. Any such errors registered must be deemed possible and pass defaults in 3D space & time. By such means, the new rules and tools enable all geoscience disciplines to QC-QA that part in, the whole which they are expert in, is valid, under the requirement that what they approve does not make that which other experts approve, be collectively improbable.

Table G-HRQC

fits one geo & thus one petro model.

After the work described in the co-pending application referred to above, initial estimates of lithology & velocity & porosity, optionally density may be forecast for later use of velocity & density in working acoustic impedance. Once the preceding work provides a high resolution sequence based materialisation of velocity Vint = Vo + Kn +/- Vint of Ka, with depth differences between local cells and control generic +/- local well adjustments, then work starts at the resolution of the seismic trace. This requires use of a time averaging equation so that the know thicknesses per sequence per trace continue to add up to proper thickness, once treated as a packaged sum of sub-sequences.

Specific examples of the invention will now be described by way of example with reference to following figures of the accompanying drawings in which figures show the material used and the order of use:-

Figure 4a, to process geological cause controls, evolving parameters, via input material and direct sampled material.

Figure la, to process reserve and risk information, given Fig 2a input

Table 4 summarises the general tasks and workflow, luding optional actions.

Fuzzy 3D 6# # ^♦ Sequence geometry, Properties time, 2D, 3D,

7 34 velocities & depth conversion

Relate Attributes Sub 35 Resolution indicator

/ sequence S. 7a

[5-12 Apparent 8 36 Lithology 8. litho- geology] facies

40 Pseudo grav mag QC

ACE Balance 41 Re-iteration rebalance as necessary

12 42 Trap maps: depth & delta D, GRV

3D causal 19 deposition,

controls Lithology

20 deposition, Sediment

Environment

21 deposition,

structural setting

22 deposition, litho- facies

23 burial , Normal

Overburden weight

24 burial, Overburden density adjusted

25 buria1 , Overburden age adjusted

E1P1 risk 0.5879 26 burial, non vertical

3 stresses

E1P1 waste 41.207 27 burial , structural

2 inversion s/w

Systematised De-compounding Burial Alterations of Lithofacies

It is possible to derive evidence at sequence resolution using seismic interpretation data, and well data as available, for each cause you can think of. For each cause of relative increase / decrease in normal compaction, that you can think of, & more, there probably exist examples in sedimentary basins. QC that each type of change registered (in a sequence hi res grid) are possible and pass defaults in 3D space & time. The order of operation is defined above. Each of the items marked # or 'new' below acts as a filter &/or algorithm set, that works cell data mostly shapes and attributes, to derive output material of the shape, and value of the burial change anomaly. Default systems may be set with minimum, maximum, rate of change per unit distance lateral and vertical, based on the evolving model and iterations.

Note that modern seismic data and methods of working it, tend to provide accuracy of time, velocity and depth in most project space of >96%. Therefore, this material to a large extent already contains the net sum of evidence of all local normal & abnormal compactive events. Therefore the processes of the present invention to separate the normal from the abnormal, such that lithology can be sensibly quantified, with proper understanding and approval of all team experts of the velocity/ shape field, and how components of it are sum. Geofactor analysis

Data In

Optional scaling of generic to local well data.

Burial Changes by algorithms, for burial, DA normalized depth analysis

WANDA variable water bottom, & datum new

SINDA s/w structural inversion, e.g. salt # amended 1

CONDA non vertical stress compaction # 1

FANDA fault influences # 1

BINDA 1/w structural inversion, basin # 1

TENDA thermal & igneous, conductivity # 1

PENDA pressure # 1

ANDA age adjusted New

FENDA fluid, oil/ gas adjusted New 1

SLENDA slope, extension new 1

NENDA net Burial Change new 1

COLINDA calibrated lithology new 1

# converted from normalized velocity difference to normalized depth difference, using Vint & depth, on which plots Vo & Kn +/- Vint of Ka, see figure abnormal Vint & depth difference relationship.

Amended 1 adds depth (s) and age order, at which all particular causes/ effects of Vint of Ka occurred.

Wanda: water normalized depth algorithm

Wanda processing adjusts sequences beneath a surface water layer, to adjust differences between water and rock weight. Run 1 assumes water density & corrects apparent depth equivalent of underlying sequences, by the difference between this and an approximate sediment density. Run 2 is an iteration using rock densities based on litho-facies , & present density, based on velocity and density, derived using detail of porosity and fluids. It works from surface down through sequences, to input detail of vertical compaction.

• Per seismic sample or cell, water column thickness (depth) is known as 2 way time /2, times water velocity of say 1474m/s. e.g. 1 second one way time of water is 1474m depth.

• The weight of water differs from weight of water filled sediment, by rock volume at about 2.6sg or per cc, + pore volume water, a little over lsg. As porosity reduces, weight difference compared with sea water increases. So, moving from shallow shelf to slope, water depth increases, and the weight acting to compact sediments relatively reduces.

• Wanda per sample or cell, per sequence from top down, works out the relative delta depth compared to generic norm, as a function of compactive weight differences.

Anda: Age normalized depth algorithm.

Fenda, adjusts rock volume s of sediments forming traps, to replace water or brine filled pore space, with oil &/ or gas, treating oil/ gas fluids as a change in the balance /cell, between Vint = Vo + Kn +/- Vint of Ka. It slightly has the effect of subtracting an abnormal burial delta depth. Since Vint is usually good, then a change in Vint of KA usually requires changes in the relatively fixed interrelationship of Vo + Kn. It is pretty much irrelevant, except where oil/ gas is in the system, i.e. in traps or source rocks. Where work without Fenda use suggests a source rock e.g. some shale, then it may be economically useful to use it with a range of parameters, to find out how it changes Vo lithology.

Oil / gas fluid differences relative to brine are data based for access to the Fenda operation, per lithology, per unit depth, variable according to hydrocarbon type, density, saturation, and net probable effect on rock velocity.

Slenda

Slenda calculates delta depth (difference) between project sequence volumes, and generic normal behavior, for slope sediments, where structural dip post deposition is associated with extension and probably thinning. Here vertical compaction is reduced by vector systems that transfer some of the vertical weight, down dip's lesser resistance.

• Access from dip data in degrees, per sequence trace or cell.

• This is best derived from 3D data, because 2D vertical x- sections may have dip out of the plane of the section unless migrated.

• Access from the current depth and Vint values per trace sample, or optionally per trace, sequence mid-point depth & sequence Vint. At the vertical resolution required, access these depth/ Vint values to the generic norm material to estimate what lithology and its Kn would be, if geology was 'normal'.

• See Fig 3-3 wanva & slonda /slenda, where the intersect 1, gives a lithology Vo & Kn.

• Need is to know the depth difference between actual cell/ sample & 'norm' . This is generated by using dip angle in degrees times thickness of sediments at that average dip, times a scale factor, to subtract depth equivalent to tie generic normal equivalent properties i.e. Vint/poro- perms .

Nenda

Net burial change relative to generic normal is constructed, as a depth of burial difference, according to rules.

• Need is to quantify the most compaction that a sediment has undergone.

• How deep was it at maximum depth of burial?

• How much non vertical stress did it receive, and at what depth?

• Has it been uplifted, to a shallower depth, relative to present depth?

• If a sediment was once buried 500m deeper than now, and when it was shallower than now by 1500m it received non vertical stress compaction equivalent to 1000m extra burial, then the present normalized depth should be that of maximum compaction which is present + 500m, because the compaction ' CONDA' effect was later exceeded i.e. overwritten.

• The rules access tables of such effects as relative depth differences to include reference to the sequence number or name being deposited, when the older sequence was so affected.

• The rules then quantify which effects per sequence sample/ cell were overwritten in the material, and which constitute maximum effective compaction/ digenesis. This required the amended 1 material

• If a 160 million year old sediment received extra compaction say 500m equivalent by thermal, igneous cause of a dyke or sill, 60 million years ago, then got buried 1800m since, the 60 million year old event would be very largely overwritten.

Col nda

Calibration of lithology is a set of processes, worked, after delta D is generated, to double check that resultant, reconstituted lithology material is fit for purpose.

Default tables are used to highlight lateral and vertical changes within sequences and across sequence boundaries, on the basis of 'normal' associations of lithologies.

Tenda is modified to include use of rock conductivity information, accessed from tables, public domain. Run 1. Compaction characteristics (Kn & Ka) are used in velocity depth domain terms to represent digenesis of physical and chemical alterations, adjusted per cell per sequence, to tune it for thermal controls. A 3D cell volume map of conductivity coefficients is used, derived from tables of rock conductivity, and use of evolving information of rock type. To this is added information on sources of heat in the mapped volume, from basement, (low frequency correction, manual or grav-mag based) or from igneous matter as intrusions or dykes or sills moving via zones of weakness, deduced from the data as anomalies in local velocity. Run 2 uses conductivity material corrected via density evidence based on porosity and fluids, i.e. re-input for iteration.

Local well tie alterations -work module, is a term used for an alternative process of modelling net burial alteration, via processes using velocity, depth, lithology cross plots, of material before and after filter use. It started off as arbitrary, and became automated to this specification. Tables compiled from public domain material listing of local or generic world-wide averages of lithology types for sediments deposited, in basins of the structural burial systems consistent with that locally determined. The system then postulates a bulk or local well gridded data correction needed to shift sequences from present velocity depth lithology chart position to non burial anomalous position, using the constraint that deeper sequences necessarily have equal or greater anomaly potential for correction than shallower ones. This method deduces one model of burial alteration.

The method above, other than 'arbitrary', produces a model of burial alteration, by a different model. Work by filters may, in iteration one, materialize >75% of the evidence available. A primary difficulty in use of the filters, IPR1, was inability to quantify extent of cause control exerted by un-quantified sediment removed from the system, in episodes of erosion, at unconformities.

BINDA works effectively, in inverted volumes, of little erosion loss. Best practice needs to combine material from various models. The arbitrary system tends to provide missing evidence .

After the Nenda, quantification of depth difference between local trace samples and the generic data, then the local is rescaled to allow lithology to be quantified.

De-compound Depositional system.

Litho-facies . Function of lithology, environment, structural setting, climate, all to be defined per geo-cell. Derive evidence at sequence resolution using seismic interpretation data, geometry & seis strat, and well data as available. Commence by subtracting burial alterations from apparent geology. Map Lithology. Map depositional environment, and structural setting. Add palaeo climate data if available. QC that changes registered are possible and pass defaults in 3D space & time.

(WB2_DC) Deposition.

These are amended to include output from SI-2 programs, plus

(WB2_DC_CO) Chronographic Offset *

*

(WB2_DC_BP) Boundary Pattern * * (WB2 DC IC) Internal Character *

*

(WB2_DC_E V) Environment * * enter calculated lithology environment,

structural setting, palaeo climate.

(WB2 DC PI) COLINVA calibration of net litho-facies *

* * De~compound Resolution error system.

Map Apparent geology minus Burial changes and minus depo- litho- facies, to equal resolution error. QC that errors registered are possible and pass defaults in 3D space & time.

(WB2_RE) Resolution Error. *

RE = AG - (D + BC) so AG- RE = D + BC *

*

= something approaching RG real geology

( B2_RIS ) Risk Analysis * *

(WB2_OUT) Data Out *

The working of material this far allows understanding of a cause control model of geologic behaviour as seen via a domain of velocity, depth burial changes to lithology of deposition.

Means provided.

Processing run 1.

Depositional litho- facies .

The system models litho-facies as follows

□ Pp2/02 (Sl-2 processes)

□ 1. Well data as available + apparent lithology from velocity depth & compaction modeling at sequence resolution

□ 2a iteration 1,

□ 2b iteration 1

□ 2a 2

Q 2b 2

The present invention extends filters capacity to isolate information on specific types of causes of burial alteration, and are new ways to quantify net burial change cause and effect. The invention extends the range of filters used to isolate information on specific types of causes of depositional litho-facies, and new ways to quantify net litho- facies cause and effect.

The present invention uses specific filters set as defaults by processing of the input data into both burial and depositional cause control models. These models are used to remove information of high risk, and classify all information in terms of risk, relative to model. QA- QC services, & De-compound Resolution error system.

This allows checking of G&G work done conventionally, without these methods .

• Plot seismic as lithology and Vint, & Vo + K, from seismic and well data conventionally assembled.

• Plot seismic as the same, via these methods.

• Subtract one from the other, to isolate differences.

• Explain differences, by relative capacity to justify each, and within each, the extent by which all are or are not a function of one another.

This maps separately Vint as the sum of lithology and burial changes +/- any unacceptable resolution component, as error. QC that errors registered are possible and pass defaults in 3D space & time. This allows all disciplines to quantify and sort out any situations where the preferred material in one discipline is an improbable to impossible conversion of material generated in other disciplinary areas.

Capacity is put in place to allow re- iteration to rework material that includes density, (quantified after porosity is quantified) for re quantif cation using velocity & density of acoustic impedance.

Processing run 2 Cause Controls

Above listed cause control programs include capacity to rerun actions to include the causal controls exerted by density, which cannot be determined in rock volume terms, until porosity is known. In practice, the Vint recorded & used before separating out density, has in large part separated out gross rock density within the n & Vint of Ka work performed here, so big changes to AI thence Vint are not likely.

Seismic is the primary input to this work system, and seismic is based on measurement of acoustic impedance changes

AI, combined with reflected wave- form. AI is about 4 parts velocity and 1 part density [changes] .

The prior art tends to work in a velocity-compaction- depth- lithology domain, without properly quantifying compaction, so putting contingent risks across the whole work domain. The present invention works in a velocity, density, compaction and lithology, depth normalised domain. It leads to use of seismic resolved at the sample rate of say 2 milliseconds per trace, commonly better than lOmetres vertical .

Causal Control System.

Test such systems and default settings, to generate at a facsimile of seismic sequence and seismic trace definition, where such constituents of the compound are constrained to make collective compounded reason, relative to default systems .

Risk Analysis System.

Quality control

The present invention includes work and assembly formatted for statistical analysis of fit.

Risk Analysis

Fit between output data w/b 1 & output data w/b 2 is derived as a measure of probability of correct prognosis.

Working geology at seismic trace sample resolution.

This occurs once all traces at each sequence mid-point have material defining normalised maximum depth of burial Then, the AI changes in the trace are converted to velocity Vint changes, used with sequence boundary time information, and calibrated with seed point material as available, including corrections to redo & recheck depth conversion. Then, per sequence, the' time averaged' corrected thickness is reconstituted such that the sum of Vint/ time pairs in the sequence sum depth to that worked to have effected depth conversion.

Then, knowing per sample, time, Vint, Vo + Kn +/- Vint of Ka, +/- depth difference to generic normal, also made material are lithologies.

The method of the present invention also includes the provision of information on • Thermal conductivity of sediments blanketing the potential source rocks from above, and influencing heat supply from below.

• Whether or not particular sediment of particular age and geographical location at time of deposition enjoyed warm, wet local environment, +/- much organic matter.

• Source potential of shale, and other low energy sediments .

• Thermal gradients and pressure gradients, at generic normal and local depths .

There is a need to output, maybe separately, project data to be made cellular, to quantify form & properties of seals, sources, carrier beds & potential traps. The component materials are, per seismic sample: -

• A generic normal lithology quantification in depth of the velocity Vint, Vo + K, porosity, permeability, fluids, temp & pressure +/- etc.

• As may be adjusted in QC steps, a current Vint, time, depth set of 2D &/or 3D fields, plus means to compare/ contrast Poro-perm output in space with seismic pattern (seis-strat) shapes and character (attributes) .

• A local high resolution quantification in depth (& time) of trace sample lithology, and depth difference to achieve equivalent burial/ compaction depth with generic normal data. Vint per sample is split into Vo +Kn +/- Vint of KA. The latter is confirmed as a delta d, depth difference. Means are included to compare/ contrast Poro- perm output in space with lithology, and the normal and abnormal compaction.

Seismic & seismic geology is converted into poro-perms via knowledge of the difference in depth between local spatial model and the generic normal model. Each trace entered is a pseudo composite log, displaying 2-4D seismic in time or depth, as AI, Vint, Vo, K normal, (lithology) Vint of K abnormal, and as Depth difference to generic model.

The production line produces extras, as porosity, permeability, seal & reservoir via X-plot. Default systems are optionally used to ensure multidisciplinary sense at and between every sample & cell.

Primary deliverable, for normalised G&G processing of seismic, is petrophysical poro-perms, plus associated quantification of seal, reservoir and carrier bed properties.

"According to the present invention, there is provided a processor equipped for receipt of the various types of input material, specifically including output in which is generated the material parameters, per layer and per cell and assemblages of cell and layer parameters to isolate quantify and quality control the petrophysics needed to quantify the geometry of sealing rocks, so traps can be quantified.

Lithology, velocity & depth are made by the depth normalisation method per seismic trace sample, and entered. This is converted to Porosity total (& optionally, effective) then Permeability, from depth normalised charts, relationships and algorithms. Then, from cross plots of poro-perms, Seal, reservoir rock potential and carrier bed characteristics are materialised. Optionally, conversion to Pseudo Gravity & Magnetic fields, & fit to real data is performed.

An A-C-E balance may be used to constrain values (minima, maximums, gradients) to represent what is possible for the parameters concerned, under the causal controls concerned, recording adjustments made. This derives and records differences between generic normal defined values, project well derived values, and the values generated as described above. This allows reiteration to QC/QA values that do not integrate across the multidisciplinary workflow, in which each discipline sets the defaults. Then, they are made aware of cells where their default set parameters are exceeded by work which may not exceed defaults set by other disciplines. Since Vint must equal Vo + Kn +/- Vint of Ka, then reject by one discipline must make all disciplines agree a most probable A~C~E balance.

Specific examples of the invention are illustrated by way of example with reference to the accompanying drawings in which figures show the material used and the order of use. Output of porosity is of both total & effective, to avoid confusion, by use of shale component, where 100% shale has effective porosity of 0%, and 25% shale reduces total porosity by 25%, etc., either proportionately or scaled. Further, means allow performance of need to spot check, or to generate visibility of this material at or from individual data points, in seismic interpretation.

Conversion to Porosity: total - then effective.

Definition of porosity via charts was suggested in IPR1, Becvem & Spiral methods, without specification of how such charts be generated, except that they should work, by extending the methods widely used in working electric log data to the seismic method. The inventor has found that methods specified as examples in attachment works effectively in thick sequences of one homogenous litho-facies, but does not, where inter-beds occur of different litho-facies .

Methods of generating from seismic a pseudo sonic log allow lithology and depth at seismic resolution, providing means to make material DIRK concerning potential for such inter-beds to occur, and each their litho-facies and porosity, subject to use of these x disciplinary depth normalisation processes .

Figure 4:5 shows a cross plot of porosity % & permeability milledarcies log scale, upon which >15 lithofacies are commonly seismically resolvable to enable use of plots between clean sand and shale, and a few (salt, some evaporites, dykes, sills etc) plot in the lower left corner.

Conversion to Permeability.

Definition of permeability via charts was suggested by Becvem & Spiral, without specification of how such charts be generated, except that they should work, by extending the methods widely used in working electric log data to the seismic method. The inventor has found that methods specified as examples work effectively in thick sequences of one homogenous litho-facies, but does not, where inter-beds occur of different litho-facies.

Methods now produce material of permeability via a plot of pre-defined porosity, say 0 to 45%, on one axis and on the other, permeability, milledarcies, log scale. Within this structure, lines are emplaced, for each litho-facies. Permeability is derived from the intersection of the litho- facies line with porosity. As seismic resolution improves, potential exists to add lithofacies lines to the primary 15 or so most important sedimentary rock types

Figure 4:6 shows how seismic trace samples can be displayed as Seal, carrier bed, reservoir quality. Once lithology & poro-perms are quantified / sample, not only can seals & reservoir and rocks be characterised. The material evolved by the method of the present invention can be converted into Pseudo Gravity & Magnetic fields, & fit to real data .

ACE balance. Constrain values (minima, maximums, gradients) to represent what is possible for the parameters concerned, under the causal controls concerned, recording adjustments made. Derive and record differences between depth normalised values and well derived values. Iterate integration as required.

The systems include

• Specific filters set as defaults by processing of the input data into rock property effect models . These models are used to remove information of high risk, and classify all information in terms of risk, relative to model. Optionally, data can be infilled by gridding or kriging to estimate possible geological causes and petrophysical effects in any data reject volumes.

• Derivation of density as a product of litho-facies , and burial history and total porosity.

• Backing out of density & velocity to confirm correlation to acoustic impedance.

• Best fit seismic reflection 2D, 3D data Best fit AVO, inversion Grav-mag. Material can be processed to display models of seismic sequence pseudo gravity and magnetic response to enable interactive best fit between the actual and pseudo information. As follows

• Total porosity, litho-facies from above output, + associated

Charts to give density, & magnetic response.

• Sequence pseudo gravity maps are made using total porosity & litho-facies material + density conversion charted + scaling factor.

• Sequence pseudo magnetic maps are made using total porosity

& litho-facies material + magnetic response conversion charted + scaling factor. 12000384

31

• For Quality control, this invention ensures that work is assembled and formatted for statistical analysis of fit, or other purposes .

• Risk Analysis, includes Fit between output data is derived as a measure of probability of correct prognosis.

Differences from other methods and means

Seismic data is displayed in colours to represent depositional lithology, total porosity, permeability and seal and reservoir characteristics, adjusted to conform with influences identified in all parts of the present invention.

Resolution achieved is that of the seismic, i.e. +/- several metres vertical and horizontal. The processing is in two parts .

Acoustic impedance data is used, integrating it with mapped sequence velocity and density information, via both IPR pubic domain elements based on Lindseth's 'Seislog', & Oldenburg, [reverse engineering of the way synthetic seismic is made from electric log data] & new methods and means.

Specific filters can be set as defaults by processing of the input data into rock property effect models. These models are used to remove information of high risk, and classify all information in terms of risk, relative to model.

In Fig 4:7 the four columns on the right show seismic samples as total & effective porosity, permeability and seal- reservoir quality. Seismic traces are converted into pseudo sonic logs, in part by calibration with sequence depth conversion velocities, and reverse-engineering public domain methods for turning sonic data into a synthetic seismogram.

This is then converted into lithology, via the normalisation steps above, which in effect break down attributes and effects into their cause controls within deposition, burial change and resolution error. This is then converted to porosity, permeability, seal and reservoir character as above. This is all worked as seg y files, such that seismic can be painted as above.

Working seismic sample A~C~E, Attributes, Causal Geology,

Effects petrophysics, into 3D cells. This is required to allow machine quantification of gross rock volumes that are structurally or stratigraphically trapped, i.e. vertically bounded by sealing rocks.

Prom Fig 4.· 9 one can see that there is a need to auto define, from a lot of 2D horizontal or vertical seismic displays, the gross rock volume of all both types of traps. Assume the pink rock has fair reservoir and the upper / laterally bounding white rock seals . Convert 2D samples to 3D cells .

The invention accesses trap gross rock volume by putting into a new material store/ data base, just copies of data from cells that act as seal.

Where the basal area of any seal exceeds a default given size, e.g. 250m square, action is as follows.

• The highest cell is identified.

• The area of a horizontal flat surface at the base of the highest sub seal cell is checked to confirm or deny that the flat lower surface does not extend outside the area of the base of the seal.

• If it does not, the base of the next lower cell is similarly worked, continuing downwards, until the process quantifies a spill point, where the flat surface extends outside the seal base.

• Optionally, the flat surface used and moved down, cell by cell, can be tilted, if it is believed that water flow under the trap causes such tilt.

• When a spill point is defined, the trap's gross rock volume equals approximately the sum of cells e.g. default 10m thick, so 10m thick flat surfaces, inclusive, between the surface base above spilling surface and the base of seal .

Within any trapped gross rock volume that exceeds a given default volume, cell material is transferred from the previous material generation, just into the trap gross rock volume, for further work to quantify net rock volume, reserves, recovery etc., after remodelling, treating non brine pore fluids as being 'abnormal burial changes' . Before one can forecast the type of trapped fluid, some source evaluation is required.

Source evaluation.

The present invention accesses potential source rock volume by putting into a new material store/ data base, just copies of data from cells that have potential to act as source. So, a set of material is constructed in a data base, keyed to generic normal geology and depth, per lithofacies / lithology deemed to have source potential. The material so transferred to the source study volume details per cell, depth difference between generic norm and local.

The method of the present invention also includes the provision of information on

• Thermal conductivity of sediments blanketing the potential source rocks from above, and influencing heat supply from below.

• Whether or not particular sediment of particular age and geographical location at time of deposition enjoyed warm, wet local environment, +/- much organic matter.

• Source potential of shale, and other low energy sediments .

• Thermal gradients and pressure gradients, at generic normal and local depths .

From the above +/- local well data, experts in the art know how to forecast oil/ gas type, volume in place and migrated out. Of hydrocarbon sourced locally, some may then migrate upwards via carrier beds to traps .