rhove eage 201703021

RhoVeTM MethodA New Empirical Pore Pressure Transform

This presentation and all intellectual property discussed in this presentation are the property of GCS Solutions, Inc. and/or Matt Czerniak.

The Rhove method represents the culmination of a 3 to 4 year effort to combine and integrate many of the advanced, state-of-the-art concepts that are useful in today’s pore pressure estimation and theory.

…and whether you’re a driller, geologist or subject matter expert, the Rhove method presents an intuitive, user-friendly application for pore pressure estimation…


OVERVIEWINTRODUCTORY DEMO

PREVIOUS WORK - THEORY

Mechanical vs Chemical Compaction Smectite – Illite Conversion

RHOVE METHOD

Summary Virtual Model Alpha – A-Term Shale Discrimination

SUMMARY DEMOS

Untethered Mode Tethered Mode

WELL EXAMPLES using RHOVE METHOD

ADVANTAGES of RHOVE Method

With a simple mouse wheel rotation, shale discriminated sonic and density data are transformed to common estimates of effective stress and pore pressure, where convergence of the two properties offers a robust solution…

Using the density data also provides a rationale for subdivision of major flow units and boundaries (shown here as P1 – P5), which can be utilized in basin modeling simulations,…and providing an opportunity to finally integrate velocity-based methods with basin modeling approaches.


0.0

/1.08

0.1

/1.08

0.2

/1.08

0.3

/1.08

DTCO Sonic

Rhob Density

0.4

/1.08

DTCO Sonic

Rhob Density

0.5

/1.08

DTCO Sonic

Rhob Density

0.6

/1.08

0.7

/1.08




RHOVE METHOD


SUMMARY DEMOS




Many of the methods in practical use today were consolidated, cataloged and ranked through a joint industry project called DEA 119 around the turn of the Millennium. The DEA study concluded that most methods fall into one of three categories. Both vertical methods, such as Equivalent Depth, and Horizontal methods, like Eaton require fitting of a THEORETICAL normal compaction trend in depth.

Other methods would include Bowers Method, which goes directly to the velocity – effective stress relationships. The Rhove Method would also fit into the category of “other” methods. Equivalent Depth, Eaton Method and Bowers would all be examples of Mechanical Compaction approaches…


This figure (modified from Katahara, 2003) - contrasts Mechanical versus Chemical compaction approaches. In principle, fine-grained sediments compact predictably in the presence of an increasing load – However, the sediments can become unloaded and follow a separate unloading limb. In terms of sonic versus density cross plot space, having a velocity regression without a concomitant change in the density is a Classic signature of unloading.

The Chemical compaction crowd would argue that many of these same observations can be explained through clay diagenesis, primarily through smectite to illite conversion. The transformation involves the liberation of bound water (through cation exchange). The liberated bound water and the effect that any compositional changes may have on effective stress and pore pressure are the subject of debate.


Modified after Katahara, 2003 OTC

Mechanical vs. Chemical

A lot of this debate is fueled by observations such as Dutta’s 2002 Leading Edge paper that included a review of the DOE #1 well in the Central Gulf Coast region. Dutta fit a model (in red) to the XRD work of Fried (1979), which depicts the transition of smectite-to-illite-dominated clay over a very narrow depth and temperature range.

In sonic versus density cross-plot space, the data tended to cluster along two distinct trends that Dutta named EODIAGENESIS and TELODIAGENSIS, or smectite and illite.


Dutta 2002 TLE

Shallow

Deep

AllSmectite

All Illite

Dutta

after Dutta, 2002 TLE

Dutta’s furtherance of Chemical Compaction, relies on the use of the Arrhenuis Law – which includes temperature in describing the physics behind the conversion of smectite to illite.


Arrhenius Law

ki = Ai e-Ei RT

Describes the controls of temperature and time on the rate and extent of chemical reaction (Roaldset et al., 1998).

**note: subscript i denotes a parallel reaction

Following Dutta’s work, in 2003 Alberty-McLean published two separate linear trends (in sonic-density cross plot space), representing smectite and illite.

Later, in an 2011 SPE Distinguished Lecturer series, Alberty published separate smectite illite trends in velocity – effective stress space with illite having the lower effective stress for a given velocity – again through a diagenetic process, with illite having the lower effective stress for a given velocity.

Alberty had essentially gutted Dutta-Wendt’s thermal contours in velocity – effective stress - (which if you recall, temperature is a critical component in Arrhenius Law) – and embraced a more empirical approach by calling Dutta’s coldest contour Smectite, and the hottest one Illite.


Alberty-McLean, OTC 2003

ILLITE

SMECTITE

Alberty

after Alberty, SPE DL Series, 2011


ILLITE

SMECTITE

Temperature Bands (deg F) 50

100150

200250

300350

400

Dutta-Wendt

Alberty


Significantly, Alberty had moved outside the realm of Arrhenius theory and took on a more pragmatic empirical approach that included the effects of compositional change(s).

Understanding and challenging Alberty’s empirical approach became a major influence over the next several years - leading up to the formulation of the RhoVe method.



ILLITE

SMECTITE

Alberty


More recently, in 2015 Sargent et al., proposed a model for compositional changes - where they renamed Dutta’s EODIAGENESIS and TELODIAGENESIS trends (in sonic vs. density cross plot space) - MECHANICAL and CHEMICAL COMPACTION trends or limits – where the latter is consistent with Bower’s published UPPER BOUND TREND for Shales, otherwise known as his Gulf of Mexico “SLOW” trend.

Once transitioned, the sediments can continue along the Chemical Compaction trend without unloading, or move to the right and become Unloaded, or some combination of the two in the ARC region above the Chemical Compaction limit.


after Sargent et al., 2015

(eodiagenesis)

(telodiagenesis)

Sargent


PREVIOUS WORK - THEORY Mechanical vs Chemical Compaction Smectite – Illite Conversion

RHOVE METHOD


SUMMARY DEMOS




RhoVeTM Method(U.S. patent pending - copyright © 2016)

Summary Interactive (and fast). Premised on a continuum of “virtual”, normally pressured synthetic rock

property relationships. Pore pressure is calculated by directly applying RhoVe-derived Velocity &

Density-Effective Stress trends. Subsalt Applications – Handles varying shale lithologies with multiple NCTs in a predictive manner. Two-parameter approach: a-term & alpha (α); includes the effects of

compositional changes (clay diagenesis) Rationale for subdivision of major flow units, which can be utilized in layer-

based basin modeling simulations. Consistent with Bower’s Method solutions for DWGoM fine-grained clastics.

Here is a look at the virtual model, which can be perceived as fairly daunting. The top half represents a series of genetically-linked, virtual rock property relationships that are used to generate the bottom half series, which are the actual applications applied to a well of interest.

Each rock property relationship has end members, which are labeled RhoVE-S and RhoVE-E, which correspond to values of 0 and 1, respectively.

RhoVE-S represents a nominal Smectite trend and RhoVE-Epsilon is an upper limit used solely for mathematical purposes.


1.00. 1.00.

1.0

0.

1.00.

1.00.

V-rho V-z rho-z P-z

V-ES rho-ES

1.00.

If we strip away the contours, we see the end members (labeled 0 and 1), on all plots.

The dotted line represents the active curve, which is linked to all other active curves (in each of the virtual rock property series). It’s important to note that all rock properties and depths are virtual and unrelated to any well being analyzed.


1.00. 1.00.

1.0

0.

1.00.

1.00.

V-rho V-z rho-z P-z

V-ES rho-ES

1.00.

The RhoVe method is fundamentally a two-parameter approach, with something called the ”a” term representing a measure of the fractional distance between end members (in sonic versus density cross plot space).

…and ALPHA, which is a calculated property of the velocity-depth compaction trend series involving the active curve. In practice, the RhoVe method is an attempt to consolidate the “theory and observations” between these two competing rock property relationships into a single, unified empirical approach to pore pressure estimation – that includes compositional changes.

The “alpha series” can be quickly adjusted with a simple mouse wheel rotation.

All subsequent“ virtual rock property” calculations cascade from these two relationships.


1.00.

αTwo

Parameter:

a , α

a

a : fractional distanceα : calculated property

1.00.

1.0

0.

V-rho V-z

0.37

…we can take the difference between any of the virtual overburden curves and hydrostatic pressure to get normal (vertical) effective stress. We can cross-plot (or cross-reference), at any virtual depth, or position along the active curve (in velocity-depth, or density-depth,) the calculated normal effective stress,

…in order to generate velocity-effective stress, or density-effective stress (for the reference point, or the entire active curve and end members).


1.00. 1.00.

1.0

0.

1.00.

1.00.

V-rho V-z rho-z P-z

V-ES rho-ES

1.00.

ESnorm

ESnormESnorm

External publications from various sources were used to calibrate the velocity-depth compaction trend “alpha” series that extends from 0.0 to 1.0. Sources included Ebrom & Heppard’s Gulf of Mexico Shelf and Onshore trend (which set the bias for the RhoVE-S nominal smectite trend (at alpha = 0.0.). Lahann’s (2002) illite trend (shown by the series of green square symbols) set the bias for the RhoVE-Epsilon end member at alpha = 1.0.

An important consideration, is that….Ebrom & Heppard’s Rest-of-the-World Compaction trend is virtually identical to an “interpretation” of Bowers default Deepwater Gulf of Mexico (velocity-effective stress) relationship, which yields a best-fit “alpha” value of 0.37.


meters Ebrom & Heppard, 2010

0.37

0. 1.0

The form of the “a”-trend series relies on Bowers published V-Rho Power Law relationship shown here with the default constants “A” and “B” for his Gulf of Mexico “slow” trend. For the RhoVE-Epsilon and RhoVE-S series, you would simply substitute these constants and coefficients into the Power Law equation.

The series of gray contours shown graphically ranging between RhoVE-S and RhoVE-Epsilon (in sonic-density cross plot space), are simply the fractional distance between the end members, contoured in increments of 0.1 or every 10%, and is given by the expression: RhoVE intermediate equals “a” times the difference between the two end members plus RhoVE-S.

Bowers Gulf of Mexico “slow” trend (blue contour) – has an “a”-term value of 0.6, and is equivalent to Bowers “upper bound trend for shale” – (referenced by Sargent).


BOWERS GOM “Slow” Trend RhoVE-ε RhoVE-S

Vo: 4790 4800 4900A: 2953 2000 4500B: 3.57 4.2 3

ρo: 1.3 1.3 1.3

V-Rho equation (Bowers, OTC 2001) :

V = V0 + A (ρ - ρo) B

γ = 2.0

0.6

RhoVE interm: a * (RhoVE-Ɛ– RhoVE-S) + RhoVE-S

Bowers GOM “slow” trend

RhoVE-S

RhoVE-Ɛ

If we do a good job of picking our shale points using gamma ray or neutron porosity versus bulk density CUTOFF CRITERIA (or a combination of the two)

– we can remove lower clay content shales – and any sands, silts and limestone, which tend to cluster - down and to the left in cross plot – (typically outside the bounds of the RhoVE end member series).

…revealing the high-grade shale data point clusters in sonic-density cross-plot space.


Here’s a few of examples for choosing the “a”-term fractional distance values. Both of these data sets are from the Gulf of Mexico. The figure on the right reveals the delimiter, as an increase in depth does not affect the value of “a”.

For the figure on the left, the shallowest values for “a” are close to 0, and then moves abruptly to the right for the deepest section - revealing the probability that what we are observing is classic unloading…


Delimited

0.03

0.15

Examples

RhoVe-S

RhoVe-S

0.

1.0

0.

1.0

0.51

0.51

Smectite Dominated

Delimited

Illitic

Smectitic0.0

0.560.75

1.080.99

Smectite Dominated

0.03

0.15

Examples

RhoVe-S

RhoVe-S

0.

1.0

0.

1.0

0.60

0.51

0.51

unloading

This next example is from the offshore Nova Scotia introductory demo. It starts out on the RhoVE-S nominal smectite trend, and increases in depth revealing compositional changes.


Examples

0.0

0.560.75

1.080.99

RhoVe-S

In this example, the P-3 flow unit can be observed where the two properties converge for an “a”-term value of 0.75.

That is also where you would take the calibrated value for “alpha” = 0.4.

The next step would be to then map the “a – alpha” pairs for successive flow units, or other wells in the area, hoping to reveal compositional changes on a regional or sub-regional basis.


0.3

/1.08

Rhob Density

DTCO Sonic

0.00.56

0.751.080.99

0.3

/1.08

Rhob Density

DTCO Sonic

0.75

0.4

/1.08

Rhob Density

DTCO Sonic

0.75

0.6

Rhob Density

DTCO Sonic

0.00.56

0.751.080.99

/1.08

0.7

Rhob Density

DTCO Sonic

0.00.56

0.751.080.99

What we would hope to find is that if we mapped “a” versus “alpha” pairs (in cross plot) for a sub-regional area (as a composite of wells and flow units), the data would fall on a straight line trend…(upper right cross plot).

However, what we do find, is that as we begin to cross plot “a” versus “alpha” pairs, like this example that includes a multi-well, multi-flow unit study in the Deepwater Gulf of Mexico, you would likely find that the “a”- alpha relationships do fall on a trend, of the form:

“a” = 2*alpha minus alpha squared

…with evidence of a delimiter (or a plateau) formed at around “a” = 0.6, and “alpha” values greater than about ~0.37, which tends to form more of a narrow band around Bowers Gulf of Mexico “slow” trend.



Vo: 4790 4800 4900A: 2953 2000 4500B: 3.57 4.2 3

ρo: 1.3 1.3 1.3


V = V0 + A (ρ - ρo) B

γ = 2.0

RhoVE-S

RhoVE-Ɛ



0.6


Vo: 4790 4800 4900A: 2953 2000 4500B: 3.57 4.2 3

ρo: 1.3 1.3 1.3


V = V0 + A (ρ - ρo) B

γ = 2.0

DWGoM Sub-Regional Study


a = 2α – α2


0.6


Vo: 4790 4800 4900A: 2953 2000 4500B: 3.57 4.2 3

ρo: 1.3 1.3 1.3


V = V0 + A (ρ - ρo) B

γ = 2.0

(0.37,0.60)

DWGoM Sub-Regional Study



0.6

a = 2α – α2

The functional relationship may be expressed in the general form:

“a” = gamma*alpha minus alpha raised to the gamma

which not only displaces the “linearly spaced” contours,

…but, now “a” can be expressed as a function of alpha (reducing the RhoVe method to a single parameter)

The gamma term offers flexibility, and should be considered basin-specific.



Vo: 4790 4800 4900A: 2953 2000 4500B: 3.57 4.2 3

ρo: 1.3 1.3 1.3


V = V0 + A (ρ - ρo) B

γ = 2.0

a = γα – αγ

RhoVE interm: f(α) * (RhoVE-Ɛ– RhoVE-S) + RhoVE-S

(0.37,0.60)Bowers GOM “slow” trend

0.6




RHOVE METHOD


SUMMARY DEMOS




The following demos are meant to summarize the two modes commonly associated with the RhoVe method.

The first is called: “untethered” mode, -where “a” and “alpha” are worked independently.

Starting off with the linear “a”-term contours.

Here (in the upper left) is an individual sonic-density data point cluster (tied to the P3 flow unit), which we will use to determine the individual “a”-term value.


0.51

0.51

0.00

P3

RhoVe interm: a * (RhoVE-Ɛ– RhoVE-S) + RhoVE-S0.0

0.60

P3

RhoVe interm: a * (RhoVE-Ɛ– RhoVE-S) + RhoVE-S0.0

Now, with the value of “a” held constant (0.6), the alpha series can then be applied as a mouse-wheel rotation, with increments set at 0.1 alpha.

You simply guide it to fit any of the common pore pressure indicators (like the mud weight history) - it’s that easy!

– the interpretation is up to you.


0.60

P3

0.0

0.1

0.60

P3

0.2

0.60

P3

0.3

0.60

P3

0.4

0.60

P3

0.5

0.60

P3

0.6

0.60

P3

0.7

0.60

P3

0.8

0.60

P3

0.9

0.60

P3

1.0

0.60

P3

0.9

0.60

P3

0.8

0.60

P3

0.7

0.60

P3

0.6

0.60

P3

0.5

0.60

P3

0.4

0.60

P3

0.37

0.60

0.60

P3

It handles pore pressure transforms (just like any other horizontal or vertical method)

(works equally well for short, or long range interpretations in depth – this example covered only a very limited depth range).


Now here’s the second version, called “tethered” mode,

-where “a” is a function of “alpha”, and includes the pressure effects associated with compositional changes.

Starting again with the linear “a”-term relationship.

As before, we will pick the value of “a”=0.6, for flow unit P3;

However, this time we will let data cluster (P3) also represent the “a”-term value for the delimiter.


0.51

0.51P3

0.0RhoVe interm: a * (RhoVE-Ɛ– RhoVE-S) + RhoVE-S

0.60

P3

0.0RhoVe interm: a * (RhoVE-Ɛ– RhoVE-S) + RhoVE-S

0.60

…and by previously cross-plotting multi-well / multi-flow unit “a-alpha” pairs. (say, for the Gulf of Mexico) we’ve already determined the expected (variable) compositional trend.

We can then displace the linear series of contours, with contours of the form: a = 2*alpha minus alpha squared

…up to and including the delimiter set at a = 0.6.


0.51

0.51

0.00

0.00

P3

(0.37,0.60)

RhoVE interm: f(α) * (RhoVE-ε – RhoVE-S) + RhoVE-S

a = 2α – α2

0.51

0.51

0.00

0.00

P3

0.00


a = 2α – α2

0.0

0.19

P3

0.0


a = 2α – α2

0.19

0.36

P3

0.0


a = 2α – α2

0.36

0.51

P3

0.0


a = 2α – α2

0.51

0.60

0.60

P3

0.0


a = 2α – α2

Again, the alpha series can be engaged with a simple mouse-wheel rotation, with increments set at 0.1 alpha.

Where the two properties (NOW) converge offers a robust solution.

-(and the potential to one-day automate the pore pressure solution).


0.0

0.60

P3

0.1

0.60

P3

0.2

0.60

P3

0.3

0.60

P3

0.37

0.60

P3

Also, when our interpretation reaches the delimiter set at 0.37 “alpha” (which calculates to an “a”-term value of 0.6),

-we have alignment of the Bowers GOM “slow” trend, with his published and well-referenced deepwater GOM velocity-effective stress relationship - both are recognized as persistent trends in the Gulf of Mexico,

Furthermore, it can be assumed that we’ve just reached the “equivalent” of Sargent’s “Chemical Compaction” limit.



Bowers DW GoM (Default)

0.37

0.60

P3



RHOVE METHOD


SUMMARY DEMOS


WELL EXAMPLES using RHOVE METHODADVANTAGES of RHOVE Method

Now for the 2 well calibration examples using the RhoVe method.

For both of these examples, I used the gamma ray log for shale discrimination (instead of neutron porosity – bulk density cutoff criteria) - which may not always be available to us (in Real-Time or logging runs). However, using the gamma ray log yields similar results.

The first calibration example is the Hess Corporation - Jack Hays #1 well, which is located in the westernmost Deepwater Gulf of Mexico.


PI526-1 Jack Hays DW Gulf of Mexico, U.S.A.

It is an interesting well - in that it is a relatively deep test (over 26,000 feet), with sedimentation that is relatively un-fettered by salt.

Beginning with the sonic-density (cross-plot) in the upper left, the gamma ray discriminated (shale) data point clusters form a very narrow band along the delimiter - although the first data cluster (P1) is slightly below (you’ll be able to see it more clear on the final “summary” slide).

The P1 - P3 flow units are projected (in depth) in the center panel – the blue data points correspond to discriminated sonic and the gray represent discriminated density (along with continuous filtered sonic and density profiles).



0.0dT rhob

0.51

The pressure plot begins as a cacophony of curves, which includes a pore pressure interpretation (based on Bowers default Gulf of Mexico velocity-effective stress trend), applied to the continuous filtered sonic log for reference. Driller’s mudweight is in black, with ECD in gold. Connection gas is shown by the solid red circular symbols.

There is an abrupt increase in the mudweight history here at the 11-7/8” casing point along with an MDT and kill mudweight in the final hole section. Connection gas is continuous throughout the lower P3 flow unit interval.

The interpretation moves from the top down.

Once the calculation begins, the velocity and density effective stress trends (for the P1-P3 intervals) will be archived graphically, as well as digitally.



0.0dT rhob

0.51


0.1

0.51


0.2

0.51

A horizontal red bar means the interpretation is frozen for the flow unit above.

At “alpha” = 0.2, the sonic and density interpretations align for flow unit P1, which is slightly less than the Bowers method (reference) interpretation.

.



0.2

0.51


0.3

0.51


0.4

0.51

For an “alpha” value of 0.4, the RhoVe interpretation is pretty close to the Bowers method (reference) interpretation (calibrated to 0.37 “alpha”) – so at 0.4 (as expected) we’re are only just slightly above,

-but we still have a large gap between the apparent pressure history and the (pore pressure) interpretation with no sign of a regression in the sonic or density data.


However, we can simply re-engage the RhoVe method velocity and density-effective stress trends to help provide a viable pore pressure solution.

The RhoVe method applications aren’t just bulk shifts, or made up trends that are used by some practitioners to just to get an answer

- they are predictive (which is an important distinction).



0.5

0.51


0.6

0.51


0.7

0.51


0.8

0.51


0.9

0.51


1.0

0.51

Here’s the summary slide.

You can see strong alignment of the (P1 – P3) data clusters along the delimiter, with each having very different velocity and density effective stress solutions.


γ= 2.0

0.51

The second RhoVe method calibration exercise involves the BP Kaskida (subsalt) discovery well (in the south-central Deepwater Gulf of Mexico).


KC292-1BP2 Kaskida DW Gulf of Mexico U.S.A.

Two data clusters, or flow units are recognized (labeled P1 and P2), which represent dominantly Oligocene and the Paleogene (Wilcox equivalent) reservoir section, respectively.


As I start the mouse-wheel rotation, the density approaches the sonic from left to right. They converge and move together once they reach the delimiter value (a=0.51)

Increasing alpha values required to match the drilling history indicators (beyond the “a-alpha” delimiter), are most-likely related to ongoing Chemical compaction, consistent with Sargent’s model.


0.0dT rhob

P1 P2

0.51

0.1

0.51

0.2

0.51

0.3

0.51

0.4

0.51

0.5

0.51

0.6

0.51

0.7

0.51

0.8

0.51

0.9

0.51

1.0

0.51

In this summary plot, you can see P1 and P2 data clusters align along the delimiter. The relatively high “alpha”/low effective stress trends are consistent with ongoing chemical compaction.


γ= 2.0

0.51



RHOVE METHOD


SUMMARY DEMOS




Advantages of the RhoVe method are that it’s so simple and interactive, one of the future products envisioned, would include an iPad app for Real-Time Operational use.

All you need to perform (pore pressure) calculations is the addition of a “virtual” mouse wheel.


Other advantages include efficiency through simplicity - you don’t have to be a subject matter expert to use it. RhoVe method is universally applicable (works as well for unconventionals as it does for conventionals, and has subsalt applications). The concepts are simple to learn and are useful for:

Prospect ExplorationProspect MaturationOperational use

Provides a rationale for subdivision of major flow units, which can be utilized in layer-based basin modeling simulations; and finally, by including compositional changes, the RhoVe method offers the potential to not only improve pre-drill and real-time forecasts, but also one day automate pore pressure solutions.


Advantages Efficiency through simplicity -

RhoVe method has universal application -

RhoVe method provides interactive solutions for: Prospect Exploration Prospect Maturation Operations

Advanced pore pressure modeling through a user-friendly application

Rationale for subdivision of major flow units, Potential to one day automate pore pressure solutions.

RhoVeTM

RhoVeR (Remote) Drone

Nader C. Dutta (Schlumberger - WGC)Jim Sabolcik (CVX)Mark Alberty (Hess)Keith Katahara (Hess)Glenn Bowers** (AMT)Phil Heppard (COP)

Jana Marjanovic (UT Geofluids*)

ConocoPhillips

Thank You!

rhove eage 201703021

Science