TECHNOLOGY Model improves oil field operating cost estimates

John L. Glaeser
Petroleum Frontiers International
Houston

A detailed operating cost model that forecasts operating cost profiles toward the end a of field's life should be constructed for testing depletion strategies and plans for major oil fields.

Developing a good understanding of future operating cost trends is important. Incorrectly forecasting the trend can result in bad decision making regarding investments and reservoir operating strategies.

Recent projects show that significant operating expense reductions can be made in the latter stages of field depletion without significantly reducing the expected ultimate recoverable reserves.

Predicting future operating cost trends is especially important for operators who are currently producing a field and must forecast the economic limit of the property. For reasons presented in this article, it is usually not correct to either assume that operating expense stays fixed in dollar terms throughout the lifetime of a field, nor is it correct to assume that operating costs stay fixed on a dollar per barrel basis.

Historical data

A useful method for comparing historical operating cost data for different fields, where the reserve size and producing rate differ widely, is to normalize the costs on a dollar per barrel basis by plotting the cost per barrel against the fraction or percentage of reserves depleted.

The percentage of reserves depleted is simply the cumulative production at a certain date, divided by the estimated ultimate recoverable reserves.

In plotting the cost histories for a number of fields, three fields had similar plots. These fields are characterized as follows:

• North American fields, where development was initiated in the 1960s or early 1970s

• Fields with ultimate recoverable reserves exceeding 200 million bbl

• Fields having a significantly long operating history, where reserves have been produced in excess of 88% of the estimated ultimate recoverable reserves

• Fields with development done in an optimized manner, maximizing the project's present worth.

Fig. 1 [18608 bytes] shows a curve fit of the historical data for the three fields. In all cases, costs are presented in constant 1996 dollars.

Although the fields examined are primarily oil producers, any gas production sold is included as barrels of oil on a BTU equivalent basis. All costs include head office overhead allocated to the operations and any property taxes, but exclude severance taxes or other production tariffs.

In addition to these three fields, a number of other fields exhibited about the same normalized cost trends.

The production profiles of these fields also followed a trend. In this case, production rate, in terms of ultimate recoverable reserves divided by annual production, is plotted against the fraction of reserves depleted (Fig. 2 [20763 bytes]).

Based on Figs. 1 and 2, the production rate and operating expense profiles of a hypothetical field with 400 million bbl of recoverable reserves was calculated, and plotted in Fig. 3 [22926 bytes]. The operating expense profile is characteristic of fields with optimized development, and shows a peak or maximum in absolute dollar terms shortly after the field goes on decline.

In this example, the operating costs peak when 78% of the reserves are depleted. The author has observed that declining oil production is the factor usually prompting the operator to institute cost savings measures intended to prolong field life.

Fig. 4 [19087 bytes] presents the same operating cost profile, as in Fig. 3 [22926 bytes], on a dollar-per-barrel basis. This figure illustrates the problem of attempting to forecast future operating cost trends on either a fixed absolute cost or fixed dollar-per-barrel basis.

If a dollar-per-barrel cost is fixed at any time during the field life, the economic limit will clearly be underestimated. In all the cases observed, dollar-per-barrel costs increase over time, with some up and down fluctuations; although, the rate of increase tends to diminish because of continued efforts by the operator to hold down costs.

In Fig. 4 [19087 bytes], the dotted line illustrates operating expense if it is assumed that the maximum costs remain fixed in absolute dollars after the peak expense is reached. If, for example, a limiting operating expense of $10/bbl is assumed, the field will reach its economic limit in 17 years under the fixed dollar operating expense assumption, rather than 23 years if cost reductions are considered.

It is often counterintuitive to expect operating expense in absolute terms to be reduced in the latter years of a field's life, but typically operators have made significant cost reductions during this period.

Operating costs are often predicted using fixed cost and variable cost components. However, what is normally assumed as a fixed cost may not in reality stay constant. For example, the cost of field personnel are often considered fixed, but experience has shown that operating manpower may be significantly reduced in the latter stages of a field's life.

Operating cost models

Although prior histories can be used as a guide, it is always preferable to develop a model for calculating future operating cost trends of any specific field. A unique model for a field will not only help in evaluating future investment economics, it can also assist in developing cost-reduction strategies.

A sufficiently detailed operating cost model may be used for a number of purposes such as:

• Forecasting the field's economic limit for a limiting well head oil price.

• Evaluating the true economics of additional investment, such as a facility expansion by estimating the true incremental operating cost associated with the incremental production.

• Evaluating the impact of various reservoir operating strategies, such as the operating expense impact of a waterflood expansion that results in increased fluid handling and oil dehydration costs.

• Evaluating the impact of cost reduction strategies, such as consolidating multiple operators, or outsourcing certain functions to contractors.

Based on a successful model previously developed, the following four steps are suggested for creating a useful operating cost model for a field:

1. Assemble a detailed base line reservoir forecast consisting of oil, gas, and water production streams, injection streams (water, gas, miscible injectant), and number of operating wells by category (producing oil, producing gas, water injection, gas and miscible injection). The forecast should include a schedule of capital expenditures and the associated facility expansions and wells drilled.

2. Based on reservoir data and existing and future planned facilities, forecast the amount of equipment or machinery in service at each time step. This may include separator trains, compressors, injection pumps, etc. Based on the equipment forecast, estimate the number of field personnel required to operate the facilities and the wells in service.

   Major cost reductions can be achieved by taking equipment no longer needed out of service, both by reducing maintenance costs, and the number of people required to operate the equipment. Good estimates of the number of people are critical, because wages and labor burden are often the largest components of total operating expense.

   Based on operating manpower, relationships can be developed for forecasting support costs associated with the manpower. These support costs can include materials/warehousing, safety, training, human relations, head office overhead, and others.

3. Analyze historical cost data, and organize the data into categories that can be used for forecasting purposes. Typical categories would include labor costs, chemical costs, well maintenance and stimulations, well workovers, logging and wire line expenditures, parts and materials, major repairs, electricity, vehicles, and transportation.

   Unfortunately, accounting data are usually not presented in discrete categories suitable for forecasting purposes, and significant reorganization of the financial reports is required.

4. Based on historical data, develop formulas for predicting future operating costs for each variable using the personnel forecast and reservoir data categories as dependent variables. For example, chemical costs may be related to water production, and well workover costs may be related to the number of wells in service.

   It is necessary to also include an allowance for unanticipated events causing property loss. The property loss costs can be included either on an average or periodic basis.

The Author