Gas from the Western Canada Sedimentary Basin - What Next?

Jennifer A. Sears and Paul L. Sears* (

*Author to whom correspondence should be addressed

 A poster presented at the 3rd International Workshop on Oil and Gas Depletion, Berlin 25th - 26th May, 2004.

Introduction

The rate of extraction of conventional gas from the Western Canada Sedimentary Basin peaked at the end of 2001, at approximately 470 million cubic metres per day. This gas supplied all but a few per cent of Canadian consumption and about 23% of total North American consumption. At peak, the basin was providing 330,000 tonnes of gas per day.

A rapid increase in gas extraction rate in 1992 - 1995 was achieved by making an average of about 4,000 new connections per year (note: one well may provide more than one connection to gas). By 2000 - 2001, however, 12,000 new connections per year were required to maintain output. In 2002 and 2003 there were production declines of 3.3% and 3.7%, in spite of a continued high rate of drilling.

Here we analyse data from Canadian National Energy Board reports, and project an increased rate of natural gas production decline until the year 2015.

Method and Results

Figure 1 depicts two scenarios for production of conventional gas from the WCSB, along with actual production from 1990 to the end of 2003.

(From the (Canadian) National Energy Board, "Canada's Energy Future - Scenarios for Supply and Demand to 2025", published 2003)

Rate of Gas Production

The average initial gas production rate for a new connection, as well as the rate at which production declines, was drawn from another National Energy Board Report ("Short-term Natural Gas Deliverability from the Western Canada Sedimentary Basin 2002-2004"). This report contains a model that describes the output from a connection as a high initial output which undergoes two exponential declines - a rapid one for one to two years, followed by a slower one. This information is provided for "average" wells for each year from 1990 to 2001. These decline rates are now considerably higher than they were at the beginning of the period. The initial decline of a new connection is now about 40% per year compared to 18% in 1990.

The output of a gas connection enters a yet slower decline after a few years. This may be seen from the relatively slow decay of output (10% per year) reported for pre-1990 connections in the latter part of the 1990's. Integration of the decay curve for an "average" connection each year provided an estimate of the total gas that will ever be drawn from the connection made in that year, i.e., an estimate of the average pool size. This was done for the years 1990 to 2002, using the first and second decay rates from the NEB short-term reports and a third 10% decay rate beginning after five years.

Figure 2 depicts the lifetime total output from an average connection made in each year, plotted against the total number of connections beginning in January, 1990. The total amount of gas per connection decreased by a factor of about four during the years 1990-2002. The National Energy Board, in its "scenarios" for the future, assumed that the behaviour of future connections for some years will be the same as those from 2001. It seems more likely, however, that the pattern of decline shown in Figure 2 will continue.

Figure 3 depicts the total gas available from connections made in 1990 or later in the Western Canada Sedimentary Basin as a function of the number of connections during and after 1990. The total gas may be estimated as the area under the curve in Figure 2. The gradient of the curve is the total gas per connection.

The curve in Figure 3 may be projected to predict the total output from the formation for future connections. This will come to a horizontal asymptote as the resource is exhausted. If this curve is regarded as a "creaming curve", it might reasonably be expected to have a hyperbolic shape, and such a curve does in fact provide a good fit. A hyperbolic curve for Figure 3:

total gas from post-1990 connections    =    c₁ - c₂/(number of connections + c₃)
(c₁, c₂, c₃ constant)

implies a curve of the form:

gas per connection    =    c₂/(number of connections + c₃)²

for Figure 2. Such a function was fitted to the data and is also shown in Figure 2. This function permits one to project the total gas for future connections. This curve was used to predict output for future years.

Figure 4 depicts estimates of total output at the beginning of each year from 2002 to 2015, along with the National Energy Board scenarios.

Decline patterns for connection output were assumed to remain consistent after 2001. This permitted the initial output of a connection to be inferred from the total production. 14,000 connections were assumed for each year after 2002. This information, with actual and inferred initial production and decline patterns for connections made after 1989 permits one to forecast output from 1990 and later connections at any date. Figures for pre-1990 connections were obtained by extrapolating information from the short-term deliverability reports. Solution gas figures for the future are taken from the NEB "scenarios".

The estimation method used in this work matches well the actual production at the beginnings of 2002, 2003 and 2004 (within 2% in each case) and matches the observed decline in natural gas production. The predicted output declines more sharply after 2004, at approximately 6% per year for the next decade. By the year 2015, the output is down to half its 2001 peak. This curve is substantially below either of the NEB scenarios.

Discussion

The major source of discrepancy between the National Energy Board scenarios and this work is the gas output assumed for future connections. Gas output per connection (initial and total) did indeed continue to decline during 2002 and 2003, and there is every reason to expect that it will continue to do so.

One of the reasons for the decline is almost certainly the continued practice by many gas drilling companies of exploiting small, low-risk shallow deposits of gas. The companies are controlled by "Income Trusts" and are very risk-averse. This makes them reluctant to undertake higher-risk operations, even though these might lead to larger discoveries of gas.

The larger exploration companies are at present tending to seek more fruitful areas outside North America. It remains to be seen at what stage the combination of higher gas prices and lower connection outputs will cause a change in these policies.

Gas production curves used in these models assume that the output from a connection is independent of that from other connections nearby. Should drilling intensification continue to the point at which this is no longer the case, the long-term decay of production will be accelerated, reducing output and total yield from each connection. This phenomenon could further accelerate the decline in production from the basin.

Conclusion

The gas output of the Western Canada Sedimentary Basin has passed its peak and is declining at three to four per cent per year. The decline rate is likely to increase soon to six per cent per year. A considerable increase in exploration and development within the WCSB will be required to change this outlook.

Unconventional gas is being tapped in WCSB, and there are forecasts of very significant coal-bed methane extraction. This will also require exploration and development effort.

This projection is significantly less than the government projections of gas production from the Western Canada Sedimentary Basin. The Canadian National Energy Board models may be unrealistically optimistic because they are predicated upon consistently higher-yielding gas well connections than were observed in 2002 and 2003. A three-phase production decline model fits within 2% of the observed natural gas production at the beginnings of the years 2002, 2003 and 2004.

home