Published by the DTI Oil & Gas Directorate for the reservoir engineering and IOR community in the UK.
Send comments on this issue and contributions for next issue to iornewsletter@senergyltd.com by 15 December 2002.

MEOR - Reservoir Engineering Design Principles

David S Hughes (david.hughes@senergyltd.com) of Senergy Ltd reports on a recent simulation study undertaken as part of the DTI's SHARP programme investigating some of the reservoir engineering design principles relevant to the application of microbial enhanced oil recovery (MEOR).

Microbial Enhanced Oil Recovery (using the BOS process developed at the University of Canberra and marketed by Live Oil) was trialed in UK sector fields (Beatrice, Ninian, Murchison) in the early 1990s.

Results were ambiguous, in part because of other operations which affected production, but also because the timing and magnitude of the expected response was unclear. Because of this “bad” experience there is a lack of enthusiasm for MEOR in the UK. An analysis of the situation by PSTI in 1996 [1] identified a need to be able use a reservoir simulator to model and design MEOR processes but that no commercial simulator was available.

It is also important that in designing and trialing MEOR that a multidisciplinary approach is taken. 

Microbiologists, reservoir engineers, oilfield chemists, facilities engineers, HSE specialists should all be involved in designing, implementing and monitoring trials.  As a contribution to regenerating an interest in MEOR among UK operators a reservoir simulator has been used to help understand some of the design principles taking a “generic” MEOR process as an example.

It is assumed that two beneficial effects result from the generic MEOR process.  Reduction in residual oil saturation to waterflooding (S_orw) which increases microscopic sweep, and a reduction in water mobility which causes flow diversion thereby increasing macroscopic sweep.  The reduction in water mobility is modelled using a resistance factor (RF) used as a multiplier on water viscosity. The initial value of RF is one.  Both of these effects were seen in a laboratory study undertaken by Statoil [3] which has led to the full scale development of the Norne field in the Norwegian sector of the North Sea using MEOR [4].

The mathematical model assumes two components (microbes and nutrient).  The microbes can adsorb onto the rock (sessile) or exist in the mobile water phase (planktonic).  Microbial growth depends on nutrient concentration, but reduces to zero as maximum sessile and planktonic populations are reached. The values of S_orw and RF depend on the amount of adsorbed microbes.  S_orw goes to zero, and RF reaches its maximum as the maximum adsorbed population is reached.  How quickly this occurs depends on growth constant.

It is recognised that this model is very simple and most likely over optimistic in the results it produces, but it allows reservoir engineering sensitivities to be explored.  The amounts of nutrient and microbes injected, the growth rate, the conversion ratio, the adsorption rate, and the maximum sessile and planktonic population densities were all estimated.  None of these numbers were based on known microbial processes or experimental results, rather numbers required to produce measurable reservoir engineering benefits were used.

In the absence of the source code of any other simulator being available to the author, the model was incorporated into the public domain black oil simulator (BOAST) developed by the US Department of Energy (DOE).  The development started from the version MTS (Microbial Transport Simulator), again developed by the DOE which already had some of the required features added [2].  The developed simulator has been renamed BOOST!

A simple model of a reservoir element was used with porosity=0.25, S_oi=0.85, and S_orw=0.22.  The model is 2500 ft long (subdivided into 25) gridblocks and 1000 ft wide (subdivided into 10 gridblocks).  The model has six layers each 10 ft thick.  The upper three layers have areal permeability of 100 mD and the lower 3 layers 1000 mD, k_v/k_h is 0.1. An injector controlled on a rate constraint of 8500 stb/d is placed in (1,1) and a producer controlled on a bottom hole pressure of 2000 psia is located in (25,1); both wells are completed in all six layers. The model can be envisaged as half of a pattern 2500 ft long by 2000 ft wide with a single injector and producer.  All results are reported for the half pattern. The oil in place is 5.7 MMstb (half pattern).

 
Figure 1: With S_orw reduction and RF effects (maximum RF=3.16) (click image for larger view)

For the base case the maximum RF is 3.16 and the growth rate constant 1 day^-1. Treatment starts at 300 days.  Figure 1 shows the oil rate, injector bottom hole pressure (left axis), watercut, and percentage oil recovery profiles (right axis) for this case.  An effect on BHP is seen more-or-less immediately, but it is a further 250 days before there is any effect on the oil rate or watercut profiles.  In this case, with continuous treatment from 300 days, the incremental recovery at 2000 days is about 16% of oil initially-in-place (OIIP).  Note, however, that the final pressure drop across the model increases from about 900 psi without the treatment to about 1700 psi with the treatment.  The final pressure drop has therefore increased by a factor of ~1.9.  This ratio is dependent on the water relative permeability at the original S_orw (0.6) and the maximum RF (3.16).  The ratio is in fact the product of these two numbers.

Figure 2: With S_orw reduction but no RF effect (click image for larger view)

An increased BHP at the injector is regarded as one of the signs that an MEOR process is working.  However, this can be problematic particularly if the maximum injection pressure is restricted, as this might reduce the rate at which water can be injected thereby detracting from any potential benefit from MEOR.  Figure 2 shows the results of a case where only the S_orw reduction effect is modelled, i.e. RF remains at its initial value of 1 throughout.  Although the timing of the incremental recovery is about the same, there is now a decrease in the pressure drop across the model.  In fact the ratio of the final pressure drops with and without the treatment is water relative permeability at original S_orw (0.6) x RF (1) = 0.6. However, the incremental recovery at 2000 days is significantly reduced (to around 7% of OIIP).

Figure 3: With S_orw reduction and RF effects (maximum RF=1.67) (click image for larger view)

Thus both an increase and a decrease in injectivity can be indicative of the MEOR process working.  If the maximum RF is set to 1.67, i.e. so that water relative permeability at original S_orw (0.6) x RF (1.67) = 1, then it would be expected that the final pressure drop would be the same both with and without the treatment.  Figure 3 shows that this is the case.  No effect is seen on pressure and it is a further 250 days before any response can be seen in oil rate or watercut.  In this case the incremental recovery is around 10% of OIIP.

If this situation prevailed in a field treatment, it would be beneficial in the sense that there is no injectivity impairment, but it would be problematic in that there would be no indication for a considerable time either at the injector or producer that the process was working.

Figures 1 to 3 are just examples of a variety of simulations run using this model to investigate the sensitivity of the MEOR process to reservoir heterogeneity, microbial growth rate, effect of RF effect only etc.  Although the model is relatively simple it has provided some useful insights into how a reservoir would be expected to respond to an MEOR treatment and further work is planned.

Simulations are useful in demonstrating expected reservoir response for use in the design, monitoring and analysis of MEOR field applications.  The results show what might be desirable characteristics of a microbial process as well as what should be avoided.  More knowledge is required of the microbial processes themselves (particularly at the reservoir scale) so that the appropriate mathematical models and data are available for use in reservoir simulations. 

More results from this study can be obtained from the author.

[1] "Reservoir Biogenics and its Application to Improved Oil Recovery" Emerging Technology Status Review, PSTI October 1996

[2] "Modelling and Laboratory Investigation of Microbial Transport Phenomena in Porous Media, M-M Chang, F T-H Chung, R S Bryant, H W Gao and T R Burchfield, IITR/NIPER, SPE 22845, October 1991

[3]"Microbial Enhanced Oil Recovery" Terje Torsvik, University of Bergen, Eimund Gilje and Egil Sunde, Statoil: 5th International Conference on MEOR & Related Technology for Solving  Environmental Problems, Plano, Texas September 1995 14.

[4] "Statoil Introduces Bacteria Method for Boosting Norne Recovery" Offshore, April 2001

Disclaimer: The material available on this website is designed to provide general information only. Whilst every effort has been made to ensure that the information provided is accurate, it does not constitute legal or other professional advice.
Please note: The Department of Trade and Industry cannot be held responsible for the contents of any pages referenced by an external link.