Saturday, October 24, 2015

Geosteering and Microseismic as Aids to Reservoir Characterization and Modeling in Source-Rock Mudstones



Geosteering and Microseismic as Aids to Reservoir Characterization and Modeling in Source-Rock Mudstones


Mudstones are vertically quite variable due to changing depositional conditions over long periods of time of slow deposition. Bedding tends to be thinner in shales and mudstones than in siltstones, sandstones, and massive limestones. Mudstones and shales are characterized by low vertical permeability and relatively high horizontal permeability due to bedding, anisotropy, and compacted horizontal microcracks. Organic porosity due to catagenesis may be developed in high TOC zones with effective permeability barriers in bounding beds. Regional near-vertical joint systems may be reactivated by hydraulic fracturing serving to connect different porosity zones to the wellbore. Joint density and spacing are factors in the effectiveness of such connections. Several of the reservoir mudstones are characterized by low clay content, high brittleness due to silica and/or brittle limestone, and interbedded limestone that may increase or decrease brittleness depending on its composition, grain size, and rock mechanics. The vertical variations include variations in anisotropy, porosity, permeability, hydrocarbon saturation, fossil content, and degree of brittleness at most scales. Thus there should be overall reservoir quality variability at most scales as well. Due to all these vertical scale variations it can be deduced that certain zones will better accommodate and propagate induced hydraulic fracturing. Certain zones will have better stratigraphic frac containment. These same certain zones will have better drill bit containment during drilling of the lateral.


Microseismic is one tool of confirming zones of better frac propagation and stratigraphic frac containment. Other methods to add and compare would involve examining frac records, comparing rock mechanics of the same zones in different wells, and noting production records. Apparently, microseismic can indicate how much of the SRV is propped. Having such knowledge would aid any comparative analysis of production from different zones. Apparently, Microseismic, Inc. does this with their ‘discrete fracture network analysis.’ They claim they can determine both where the fractures were created as well as where the proppant went. If such technologies can be proven to be reliable – there may be interpretive and data processing issues as in seismic in general – then they can be very valuable in providing the detailed geological characterization necessary for reservoir production optimization. More recently microseismic has been used to monitor induced seismicity from waste water injection wells, to monitor refracturing, and to monitor reservoir drainage events. Post-fracturing changes in the near-wellbore stress field can also be monitored with microseismic. 

Geosteering horizontal wells naturally works out the exact nature of the structural characteristics along the laterals, if correlated and interpreted correctly. This detailed structural characterization can aid reservoir modeling and well production analysis. Production irregularities between wells in the same area and in similar zones highlights the need for more detailed subsurface characterization. Targeting, geosteering efficiency loss, hydraulic fracture configuration, and proppant placement all can lead to production irregularities. Geosteering can also optimize access of the best reservoir and microseismic can find the best stratigraphic zones that frac well and so lead to frac optimization. Although microsesimic is typically a post-frac analysis technique, with infill drilling the nearby stress-fields have already been altered so in terms of infill drilling it can be a pre-drilling technique. 
 

Mapping out and characterizing the target interval in terms of bedding, anisotropy, reservoir qualities, geochemistry, and rock mechanics is typical to core analysis and ideally should precede drilling. One can also characterize the target zone by analyzing its drilling and geosteering properties. Zones that tend to confine the drill bit are likely to tend to confine the frac. Since in mudstones there can be variability at thin-bed scale it has sometimes been observed that very thin zones will have high or low rates of penetration (ROP) during drilling.  


More operators seem to be noting that they have had better success staying in the best rock for the greatest amount of the lateral, regardless of the thickness of section. Smaller target intervals have become the norm as the best rock in several plays and play areas has been identified. In the Marcellus and other plays there is so much interbedded limestone that the ideal individual shale zones best to drill can be quite thin.


When modelling a horizontally-drilled reservoir after it has been drilled one needs to determine first the stratigraphic interval contacted by the borehole which is a function of geosteering efficiency within the target interval. This would be the primary reservoir access. Next the reservoir contacted by the frac must be considered. This would be the secondary reservoir access. Most of the frac should ideally stay within the primary reservoir access interval with some connections to other stratigraphic intervals, presumably many through reactivated regional near-vertical joint sets. Mudstones, being generally more thinly bedded than most rocks and often being anisotropic, should have a tendency to have better overall frac containment in thin zones regardless of joint reactivation. I don’t know if this is true across the board but from microseismic examples I have seen they do seem to have good horizontal frac length and limited amounts of frac growth out of zone, generally speaking, and that is of course the goal. Bed-scale differences in porosity and permeability must be considered. These source-rock mudstones typically have nano-scale porosity and predictive techniques like Darcy’s law are not wholly applicable – they show that less fluid (oil and/or gas) will flow than does – so natural fractures are necessary (double-porosity model). Catagenic microfractures offer a third porosity type, the main type for some shale plays such as the Point Pleasant. Ozkan suggests there is mixed flow in shale matrixes: Darcy flow in fractures, Darcy flow in matrix micropores, and slip flow in matrix nanopores.  Induced hydraulic fracturing also rejuvenates existing natural fractures (typically calcite filled) which is important for increasing fluid flow. 
  

One might employ simple volumetrics utilizing thickness, porosity/permeability, and primary reservoir access to predict reserves. Ideally frac analysis would also be included: frac half-length, frac height, frac energy, etc. of each frac stage. One might then estimate stimulated reservoir volume (SRV). Flow characteristics are different in source-rock mudstones than in conventional reservoirs. Micro-fractures produced by thermal maturity catagenesis typically are the initial pathways of fluid migration in these rocks. Factors such as matrix permeability (extremely low in some plays), natural fracture permeability, natural fracture density, and hydraulic fracture conductivity affect fluid flow rates. Both natural fractures and induced hydraulic fractures are required to economically produce these source-rock mudstone reservoirs. The induced hydraulic fractures also reactivate, expand, and further connect the existing natural fractures.


In terms of sequence stratigraphy high gamma ray indicates flooding surfaces. Bedding and lamination planes are planes of weakness in hydraulic fracturing. One reason is that they lack authigenic cement, or in situ cement (not transported), along the contacts. Thus the thin bedding in shales and mudstones may frac favorably compared to thicker bedded rocks. Induced hydraulic fractures in vertical wells can clearly show stratigraphic zones that frack preferentially with frac records and especially with microseismic. Microseismic can also indicate seals/barriers and ductile and brittle zones. Cabarcas and Slatt, 2014, noted that strat sequences like cleaning upward or dirtying upward ones can be characterized in terms of brittleness and ductility, as “brittle-ductile couplets” the ends of which define sequence boundaries and flooding surfaces within those sequence boundaries. The boundaries between ductile and brittle in the couplets can define lower order sequences (2nd or 3rd order).

Another issue, perhaps relevant to differences between zones is "rock typing." Even if rocks are of somewhat different depositional types and grain sizes they may have undergone similar diagenetic processes and thus have similar properties relative to other rocks. Rocks that are typed as similar may have similar reservoir properties in terms of fluid storage, capillary pressure conditions, flow, and permeability. Rocks can be typed through petrophysical methods and log and core analyses. However, as long as there are no major changes in porosity, permeability, or water saturation, rocks can be considered more or less similar.  
  

The role of geosteering in optimizing reservoir output is simply primary accessing of the largest possible amount of the best rock in terms of gas content and flow capabilities. It would be nice to see a microseismic study of different stratigraphic zones of a target interval where there are perf clusters to compare values like frac half-length, growth, frac energy, and the path of each stage through time. It seems likely that in play areas like the western Marcellus play there will be thin hot shale zones that have the best qualities. Geosteering can help characterize how zones drill and that can be compared to how they frac and how they produce. It is important to delineate the best zones, optimize primary access to them through effective geosteering, determine access to them through fracking, and further compare production characteristics.  
    

References

Understanding Fluid Flow in Shale and Modeling Fractured Horizontal Wells – by Erdal Ozkan, Colorado School of Mines – SPE Distinguished Lecturer – 2009? (www.spe.org


Numerical Simulation of Gas Transport Mechanisms in Tight Shale Gas Reservoirs – by Jun Yao, etal – in Petroleum Engineering and Mechanics – Petroleum Science – December 2013, Vol. 10, Issue 4, pp 528-537


Releasing Shale-Gas Potential with Fractured Horizontal Wells – by Erdal Ozkan, Colorado School of Mines – SPE Distinguished Lecturer – 2012 (www.spe.org


Sequence Stratigraphic Principles Applied to the Analysis of Borehole Microseismic Data – by Carlos Cabarcas and Roger Slatt in Interpretation Vol. 2 No. 3 (August 2014) p 1-9


A Seismic Shift in the Subsurface, in “Completing the Well” (Drilling Contractor), Sept. 3, 2015 – by Katie Mazerov, Contributing Editor

Rock Typing: A Key Parameter in Reservoir Simulation - by Nabi Mirzaee, posted in OilPro, Feb. 2016




No comments:

Post a Comment