The Milk River Shallow Gas Pool: Role of the Updip Water Trap and Connate Water in Gas Production From the Pool

Abstract

Abstract It has long been known that the Milk River A Gas Pool is located downdip of an aquifer which outcrops in Southern Alberta. In the past numerous suggestions have been made explaining how this unconventional trap of gas below water Could exist, These suggestions were without theoretical or practical application. In this paper a theoretical model of the trapping mechanism is described and provides a clear explanation of the phenomenon. It is shown that buoyancy forces are not dominant until pore throats become larger and interfacial tension forces are less than buoyancy. A pressure-depth plot for the Milk River provides additional understanding of the hydrodynamic state of the water and gas in the pool. Application of the theoretical model also provides a basis for understanding edge wells whose production and recoveries are influenced by the aquifer. Introduction The Milk River Gas Pool in Southern Alberta and Saskatchewan is the largest non-associated gas pool in the Western Canada Basin. This shallow, low pressure, low permeability, shaly reservoir is estimated to contain original gas in place of over 218.3 10(9) m3 (7.7 Tcf) and was discovered accidentally in 1883. The Canadian Pacific Railway was drilling for water west of the City of Medicine Hat and struck gas in what is now known as the Milk River A gas pool. Gas has been used locally since the early 1900's but it was not until the 1970's that the pool received the attention of a large number of operators and has seen significant development. The Milk River A pool now covers over 1.6 million hectares (6200 square miles) with in excess of 14000 wells. The Milk River has long been identified as having an aquifer UPDIP of the gas pool and water production plays a major role in the ultimate economic recovery of reserves. Defining the limits of the pool and maximizing recovery along the aquifer edge is better accomplished with an understanding of the trapping mechanism involved. Within the gas saturated portion of the pool it is important to understand the mechanisms controlling connate water production in order to optimize depletion strategies. This paper gives an interpret at ion of these mechanisms based partially on a compilation and interpretation of previously published information about the aquifer hydrogeology, rock properties and the origin of the waters and gas in the pool. New data will support past interpretations of fresh water invasion and water movement in the aquifer. It is also shown that gas well production is dominated by the production of 1) aquifer water or 2) connate water each having distinctly different characteristics. Of significant importance to the results of the study was the need for understanding this large pool on a regional basis before applying concepts to local optimization of an individual property/well. Background Some of the largest gas accumulations in North America are found in low-permeability Cretaceous sandstones in structural locations with GAS TRAPPED DOWNDIP OF WATER. Numerous papers have been published which detail this special form of trap specific to the Elmworth Deep Basin area of Alberta. These references make a number of suggestions to explain why these unusual traps exist; 1) downdip water flow holding back gas, 2) water, blocks in narrow pore throats, 3) a lack of buoyant forces on the gas, 4) faulting, 5) facies controlled trapping and 6) combined stratigraphic-diagenetic trapping. Rice and Claypool in 1981 suggested that the trapping is due partially to biogenic gas being retained in solution in the interstial waters as well as free gas being trapped by capillary pressures in low permeability rock and possibly early diagenetic creation of carbonate cements. Gautier further elaborated on these points in a subsequent paper latter in 1981. The interpretation presented in this paper is from a reservoir engineering rather than geological perspective and is meant to enhance production rather than exploration. It will be shown that the trap is possible when the forces due to buoyancy are less than those caused by interfacial tension. Principles presented by Schowalter describing membrane or cap rock trapping are related and an excellent source of reference. Meyboom in 1960 gave the first comprehensive review of the aquifer portion of the Milk River pool, Meijer-Drees and Williams and latter Meijer-Drees and Myhr dealt with geological aspects of the gas pool. Several others have presented more information on the fresh water invasion into the aquifer. The development, reservoir performance and production from this pool is much different than that of conventional sandstone reservoirs and numerous papers published during the 1970's discussed techniques used during development. P. 371^