Wednesday, January 6, 2010

Two Types of Shale Gas?

Happy New Year!

It seems there are two types of shale gas:

Type 1: Shallow depth (few hundred to 2000m), sorption dominated, TOC critical (7 scf/ton for each 1% of TOC). Maturity important only to improve sorption capacity. May be biogenic origin or mixed origin. Mechanism and therefore estimation methods are similar to CBM.

Type 2: Deep depth (>2000 m), compression (free) gas dominated, porosity critical (20 scf/t for each 1% porosity unit) TOC less important. High maturity very important not only to improve sorption capacity, generate the gas but to reduce liquid volume which reduces sorption and lowers relative permeability. Higher pressure improves scf/ton value for the same porosity.

Shale gas evaluation requires a comprehensive model that takes into account the following: (a) a burial and thermal history model to predict maturity and porosity; (b) the Langmuir sorption model to calculate the amount of sorption gas in the organic matter; and (c) a pvt model to calculate in situ free/compression gas and dissolved gas in the residual oil. In general, the behavior of such a model looks like the following:


These curves are shale gas capacity based on 5% TOC and 1.8% VRo. The curves will also vary with pressure gradient, thermal gradient.

7 comments:

  1. I do NOT think it is a good idea to create a new "binned" (Type 1 & Type 2) nomenclature; the whole thing is continuous; you can have a lot of free gas in shales and coal seams at low PT (e.g., sweet-spots San Juan) and you can have a lot of ad and ab-sorption at high PT. (That more methane can be stored adsorbed at low PT in very high TOC rocks compared to quantities of free gas in a traditional kerogen-void porous reservoir at the same low PT, is a true; however, how things develop with increasing PT is much less straight forward.)

    Simple standard PVT for any free gas (PV=nRTZ and a Z chart (or equation) for reduced P and T) is fine and simple for dry gases.

    However, we should be extremely careful with Langmuir though. It is "OK" at low T (if you actually have saturation and some sort of generalized maturity vs adsorption capacity function) but theoretically, it may explode in your face at some nearly unconstrained T as T increase.

    Furthermore, with increasing PT you will get stronger and stronger competition between ad- and absorption (Langmuir insufficient; you need a dual model) and there is certainly a strong (initial) composition and maturity (related composition) effect. (Not much open literature about dual sorption phenomena in/on polymers but see e.g., Milewska-Duda (around 1986-1987 .. Fuel vol. 66 I think) and Exxons kerogen absorptions work).

    To summarize: Be very careful with sorption calculations; they are very poorly constrained outside the "shallow realm" and results can easily drift into fantasy land.

    Therefore the question: Do you have any real calibration for the temperature dependency of Langmuir (only independently addressing adsorption) for different kerogens and maturity or did you just pick an arbitrary constant or expression ?

    ReplyDelete
  2. 1) I agree that things are a continuum - it is a stretch to divide it into 2 types. I was trying to bring attention to folks, who might otherwise be thinking there is a silver bullet to shale gas, that it depends on PT conditions, in addition to a host of other things.

    2) I was not clear, that by Langmiur, I meant a sorption model (not just isotherm). We have used the sorption model developed by Hildenbrand et al (International Journal of Coal Geology 66 (2006) 179– 203) which includes calibration with temperature and maturity. It is surprising that the calibration actually worked well in all the data we had tried so far (three shale gas basins - some experimental data, some desorption data ) as well as some CBM data.

    3) We have our own pvt model that handles wet and dry gas - for density and GOR.

    4) The above model is build on top of a basin modeling software that handles the generation and liquid cracking process through time -

    I tried make a single point with the post and tried not to dwell on the details- but here we are.

    Cheers,

    ReplyDelete
  3. I have not read the Hildenbrand paper (just scanned the abstract on the net.. have not been in the library for a while). I am curious about what kind of data they have in the high PT range, i.e. for P > 200 bar. Have a look at e.g., Bustin & Bustin (2008) (AAPG Bulletin; v. 92; no. 1; p. 77-86). They state that they have data which may show the exact opposite of what Hildenbrand et al. are stating. I quote from the abstract: "Assuming reasonable geothermal and pressure gradients, our data indicate that the sorption capacity will generally decrease with uplift and associated exhumation, suggesting that an initially gas-saturated coal will de-sorb gas during uplift of the reservoir."

    So here we have to 100% opposite conclusions (Since Bustin*2 are discussing the reasons for under-saturation in the shallow part of the section, contradicting Hildenbrand et al., I assume they are uplifting in a PT range where Hildenbrand et al,. are predicting increased sorption capacity with decreasing PT).

    Who is right ?

    As I indicated in my first comment: theoretically there can be cross-overs regarding the P effect relative the T effect at higher P's and it can be "explosive".

    I worked with CBM for a few years in the late 90's, had access to data from all over the world: (US San Juan & Black Warrior, Columbia, Australia (Sydney basin),Poland (Silisian basin), India, China and South Africa.) and I am not able to make the call. The lab data are biased with most data at pressures only up to 200bar. (Naturally) Data from kerogen with more aliphatic polymers have isotherms that deviate increasingly and linearly from Langmuir with increasing P (probably competing absorption). Well data are incredibly difficult to interpret, are only reliable when you have significant under-saturation and are getting more and more difficult to handle with increasing temperature (the trip up the well has profound effect on the samples).

    If you have been able to match data to model when you have free gas at high PT, then I am impressed (If the match is at low PT, it does not really address the issue).

    If you could tell us more about that, I am sure the readers would be very interested in the methodology.

    When it comes to lab test, the few results I have seen at pressures between 200 and 300 bars, were all over the place. (I will only discuss samples saturated with H2O: dry samples have often cross-overs at P below 200 bar) (I left the industry 10 years ago so excuse me if I have missed the last 10 years development .)

    Some samples do show major PT cross overs from around 200 bar. Some samples show really major cross overs from around 250-260 bars. Some sample do not show any cross-over at all in the 200-300 bar range. Some samples shows dramatic cross overs when the "gas" mixture has significant CO2 (>20%).

    So from the data I have seen, we have a major ambiguity and we could never figure out (too little data and general control of the samples) what a "typical" real world picture would be.

    Continues...

    ReplyDelete
  4. What I think can be said with some certainty though is that Hildenbrand et al., are wrong when they suggest that the typical under-saturated shallow sections are under-saturated BECAUSE low PT saturation level is higher than at high PT (even if that was true).

    The reason I say that, is the pattern of under-saturation that typically occurs in these sections for coal. Khavari-Khorasani (1997: AAPG annual convention in Dallas.) in her coal-bed methane talk showed convincing data and also simulations regarding that (using shaky but reasonable partitioning between adsorbed methane and methane in solution in pore water). The pattern of under-saturation follows the mass-distribution of adsorbents in the sections, only distorted by the sand/shale ratio of non-coal in the sections; what we are seeing are best explained by diffusive decay of the concentration gradients, as the diffusion simulations reproduced very similar patterns. Hence, a typical under-saturated section is likely to represent a late waning stage; a degassed dead "petroleum system". (Strange that the diffusion man Krooss, did not follow up on that.)

    It does not matter what the slope of the saturation PT curve was in the first place; we have arrived too late. (Her simulations actually predicted release of free gas (consistent with Bustin & Bustin) initially during uplift; but she also pointed out the data supporting that was a sore point.)

    This discussion is at the very heart of the adsorbed gas prospectivity discussion. USGS are operating with huge numbers, without any real data or theory to do the regional prospectivity evaluation and resource estimates. Basin modeling could be a key tool for early assessment, but as far as I have seen, nobody is taking the quantitative risk problem serious; just huge unsubstantiated extrapolations from USGS. (If something dramatic has not happened the last 10 years, that does not show up on google searches.)

    I would love the shale gas to deliver. I do have a nightmare though; we build a million new rigs and make horizontal swiz cheese out of our gigantic shale gas areas, only to realize most of them hardly produce and all the currently unproven areas are bs. Anybody remember the 4.5 trillion barrels of oil Salym field. That should save the world to. Not shale gas, but rather shale oil from early mature Bazhenov shales. Instead even an nuclear explosion in 1979 in the field did not do anything but screwing it further. Today, Salym will show up on a map (it is huge and it reminds me of USGS shale gas maps), but not in a table of fields with significant production. Maybe USGS thinks this is the solution to get rid of the nuclear bombs of the world :-) ?
    Currently, I cannot make my mind up about what I like most with the Pickens plan; shale gas or wind-mills. At the moment, I vote for wind-mills because those are more tangible and easy to price :-) Unfortunately, Wall-street thinks the shale-gas is easier to price... but those guys believes in endless abiotic oil as well :-) so do not expect any rational actions from that direction.

    ReplyDelete
  5. For the latest news on European shale plays, visit the Realm Energy blog at http://realm-energy.com

    ReplyDelete
  6. Prodigy oil and gas is watching developments of reserves in and across the borders of these United States

    ReplyDelete
  7. Welcome to SoftwareCorner, your source for
    pecialized software, development and training Mathematica

    ReplyDelete