Copyright 2009, Blake, Cassels & Graydon LLP
Originally published in Blakes Bulletin on Energy, January 2009
Introduction
On December 23, 2008, the Energy Resources Conservation Board (the ERCB) issued its decision in an application by Hunt Oil Company of Canada, Inc. (Hunt) (represented by Blakes) pursuant to section 39(1)(a) of the Oil and Gas Conservation Act (the OGCA) to amend its enhanced recovery scheme, by adding two water injection wells to the Kleskun Beaverhill Lake A Pool (the A Pool). Hunt also applied to the ERCB, pursuant to section 33 of the OGCA, to include wells currently designated to the Puskwaskau Beaverhill Lake C Pool (the C Pool), in the A Pool. Galleon Energy Inc. (Galleon) filed an objection to the first application on the basis that the proposed injection wells could result in premature water breakthrough because of high-permeability channels in the A Pool, and that injecting water into updip wells in a reservoir where there is an oil-water contact could result in reduced sweep efficiency. Galleon is licensee of its own wells in the A Pool and also the holder of an enhanced recovery (waterflood) scheme approval in the A Pool.
The ERCB held a public hearing in Calgary, Alberta to resolve the matter. The hearing commenced on September 29, 2008 and concluded on October 2, 2008, before ERCB-appointed examiners G.W. Dilay, P.Eng. (Presiding Member), T.R. Keelan, P.Eng., and J.R. MacGillivray, P.Geol.
Background
The A Pool is a conventional oil pool that was discovered in November 2005 and is being competitively operated with two separate waterflood schemes. Galleon operates a waterflood scheme in the south-western part of the A pool, and Hunt operates a waterflood to the east of Galleon's waterflood. Water injection was commenced by both Galleon and Hunt in 2007. All of the injection is into a common aquifer associated with the A Pool. The ERCB considered the issues respecting the applications to be:
- pool delineation,
- the need for and location of additional injectors in Hunt's waterflood scheme, and
- the potential for Hunt's proposed injectors to negatively affect Galleon's producers. Hunt and Galleon agreed that additional water injection was required in the expanded A Pool, but they had different views as to where the additional injectors should be located to optimize recovery.
Hunt submitted that its existing injection well and Galleon's existing injection wells provided minimal pressure support to the A Pool through a poorly connected aquifer. To evaluate the effect of additional injector locations on pool recovery, Hunt conducted a reservoir simulation of several cases, starting with a base case of the current operations. Hunt argued that the model results supported its applied-for injector locations. Hunt proposed new water injection in the central part of the A Pool as well as near the edges of the pool. Galleon conducted its own reservoir simulations of several of Hunt's cases, as well as two of its own, and advocated a different approach than that proposed by Hunt. Galleon proposed just using injectors closer to the edges of the A Pool. In its decision, the ERCB noted a number of differences between the models used by Galleon and Hunt to support their relative positions and Galleon stated that although Hunt's model achieved a closer history match of gas-oil ratios (GORs), Galleon's model was better able to match water production.
Decision
The ERCB agreed with Hunt and Galleon that additional water injection was needed in the part of the A Pool operated by Hunt in order to optimize oil recovery. Because of the limited water injectivity of Hunt's then existing injector in one area, a significant amount of oil production was shut in. Additional water injection would allow Hunt to increase its oil production and maintain voidage replacement.
Both Hunt and Galleon used reservoir simulation to assess the merits of their proposed additional injectors. The ERCB stated that reservoir simulation is a useful tool to assess the merits of different injector locations. However, there had been very little water production in Pool A and thus very little water production data to use in the history match. The ERCB found that it was not possible to adequately test the ability of the models to predict water production. History matching was therefore limited to mainly matching the pressure and GOR data for the A Pool. The ERCB found that Hunt's model was better able to match these data than was Galleon's model. Also, Hunt made fewer changes to the geological input to its model in order to obtain the better history match. The ERCB noted that Galleon acknowledged it would have liked to revisit the geology around several of the wells where it had made changes to the reservoir properties to try to get a reasonable history match. The ERCB believed that in order to achieve a history match, Galleon's addition of transmissibility barriers in areas of the A Pool where it did not interpret faults to be present, resulted in a model that was likely not representative of the geology of the reservoir. Although Hunt and Galleon both characterized the A Pool as being heterogeneous, the examiners believed that in generating the geological models, Hunt best captured the heterogeneity through geostatistical modelling, compared to Galleon's deterministic modelling, notwithstanding Galleon's use of variable initial water saturations. Overall, the ERCB had more confidence in the predictions from Hunt's model than from Galleon's model.
In addition to the ERCB's preference for Hunt's approach to waterflooding the A Pool based on reservoir modelling predictions, the examiners had two concerns about the waterflood expansion proposed by Galleon. First, Galleon's proposal involved using injectors that would be located in lower-quality rock. This raised the concern that there could be problems with attaining adequate water injectivity. Second, Galleon's proposal involved converting all the wells in one of the sections to injectors, such that there would be no producers in the section. Any oil in that section displaced by the injectors would have to be captured by the producers in a different section and some of the oil would thus have to be swept over a long distance. The concerns about Galleon's proposed waterflood expansion added to the ERCB's conclusion that Hunt's proposed approach was to be preferred over Galleon's proposed approach.
Having concluded that Hunt's approach was preferred over Galleon's approach on a pool basis, the ERCB went on to consider whether both of Hunt's proposed wells should be approved as injectors for Hunt's waterflood scheme.
The ERCB found that Hunt's application to convert two wells to water injection would result in more injectors than producers for Hunt's approved waterflood scheme. As calculated by Hunt, the mobility ratio for the waterflood in the A Pool was 0.67. A mobility ratio of less than 1.0 indicates that the water injectivity of an injector is less than the oil productivity of a producer after filling up the gas space in the reservoir. Hence, the ERCB stated that from a mobility ratio point of view, more injectors than producers is preferred, which, in principle, supported including both the wells as injectors.
Another consideration in determining whether the second well should be approved as an injector was the problems Hunt had been having with the injectivity of its existing injector. Including the second well as an injector would assist in supplementing any further reduction in the injectivity of the existing injector. The ERCB was careful to note that the applied-for injectors may not be sufficient to waterflood all of the expanded A Pool operated by Hunt and that one or more additional water injectors may be required to expand the waterflood to the northeast. Therefore, any injectors approved needed to be compatible with a future expansion of the waterflood.
The ERCB, after finding that the proposed injectors were appropriate, went on to consider the effect the proposal would have on Galleon. Galleon had identified two concerns regarding Hunt's applied-for injectors:
- reduced sweep efficiency by injecting into updip wells in a reservoir where there is an oil-water contact and likely contact with an aquifer, and
- the possibility of premature water breakthrough because of high-permeability channels in the A Pool.
With respect to the first concern regarding reduced sweep efficiency, Galleon acknowledged at the hearing during cross-examinations that its own model study indicated that updip injection would not be as harmful as Galleon initially thought. Considering that the dip of the A Pool is only about 0.5 degrees and that both the Hunt and Galleon models did not predict an adverse effect due to updip injection, the ERCB concluded that updip injection was not a significant concern.
With respect to the concern about premature water breakthrough, the ERCB agreed with the position of Hunt that it was important to define what is meant by premature breakthrough. As indicated by Hunt, the ERCB acknowledged that water will eventually break through to producers in any waterflood, so water breaking through to a producer is not necessarily premature. The ERCB also agreed with Hunt's definition of premature water breakthrough as being that type of water breakthrough that occurs when water arrives at a producer without displacing a mobile oil bank in front of the injected water.
Galleon attempted to extend the definition of premature water breakthrough to include the case where injected water reaches a producer several years earlier than it would in an alternate scheme that has higher oil recovery; however, the ERCB viewed that to be more related to optimization of a waterflood rather than to the definition of premature water breakthrough.
Following the adoption of Hunt's definition of premature breakthrough, the ERCB found that, considering that the mobility ratio for a waterflood in the A Pool was calculated to be less than 1.0, premature water breakthrough would likely only occur if there were an extremely narrow conduit or channel connecting a water injector with an oil producer. While Galleon interpreted there to be channels in the A Pool, it did not include in its model any of the extremely narrow channels that would be expected to result in premature water breakthrough. The ERCB noted that neither the Hunt nor the Galleon models predicted premature water breakthrough as defined above in any of the cases that were run, including Galleon's runs of Hunt's cases. This was so even though Galleon's model included a high-permeability zone that could be considered to be representative of a channel, albeit not an extremely narrow one. In the absence of specific evidence that there were extremely narrow conduits or channels in the A Pool, the ERCB was not convinced that there was sufficient reason to justify denying Hunt's applied-for injectors. The application of Hunt to amend its enhanced recovery scheme to add two additional water injection wells was allowed.
Comments
This decision by the ERCB highlights a number of important issues. First, the examiners acknowledged the importance and desirability of enhanced recovery schemes. Second, the examiners spent a significant portion of their discussion analyzing the various simulation methods employed by both Hunt and Galleon and placed reliance on that simulation evidence in reaching their conclusions with respect to which recovery scheme was more desirable. Last, the ERCB adopted the definition of premature water breakthrough advanced by Hunt. The examiners recognized that not all water breakthrough is premature and that water will eventually break through in any waterflood. Premature water breakthrough is the type of water breakthrough that occurs when water arrives at a producer without displacing a mobile oil bank in front of the injected water.
The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.