Deep-well injection as a means of liquid waste disposal is, at best, a costly and tricky operation. Nevertheless, despite the inevitable difficulties which occur, it has proven itself to be reliable, environmentally sound, and economically feasible for disposing of certain wastes in certain areas.

A mathematical simulation model has been developed for predicting the operational response of a disposal-well system. From the initial design parameters and the physical operating characteristics it is possible to estimate the cost of such an operation. Additionally, sensitivity analysis experiments can be performed to assess which design parameters, operational characteristics, or formation properties have the most significant impact on the overall system response.

Application of the model to date indicates that for favorable geologic conditions the cost of injection may range upward from $0.25-$0.40 per 1,000 gal; this figure includes O&M plus capital amortization, with the initial outlay ranging upward from about $150,000. Even a ball-park cost estimate for a given injection system cannot be done until the key parameters (waste volume, well diameter, porosity, permeability, reservoir pressure, etc.) are known for that specific site.

Sufficient data are available from secondary sources to synthesize the basic characteristics of a "typical" injection well. (For this study, approximately 75 industrial disposal wells were considered.) Typical features include: (a) 90% of all wells in the U.S. are less than 6,000 ft deep with half between 2,600 and 4,200 ft, (b) only 10% operate with casinghead pressures greater than 1,050 psi, but 50% operate between 175 and 550 psi. These and other statistical characteristics were combined to create a set of fictitious--but representative--wells. It was upon this set of "standard" or "typical" wells that the following sensitivity experiments were performed. For our "typical well" designed to operate at 1,000 psi, an increase in wellhead pressure of 50% can be expected to raise the t tal unit cost from $0.24 to $0.32 per 1,000 gal. For a given flow rate, friction losses decrease rapidly as well diameter increases. For our well, an increase in diameter from 4 to 5 in., reduces the ratio of the pressure drop to driving pressure by 57%, thus substantially reducing energy requirements as a trade-off for a more expensive well. Responses to flow rate can be evaluated. For one of our "standard" wells, an increase in the flow rate from 400 to 600 gal/day increases the initial cost of $224,000 by 53.5%, but lowered the unit cost by 21.2% from $38.2 to $30.1 per 1,000 gal. Formation impact can likewise be assessed. For our example, an unexpected drop in permeability from 60 to 40 md would increase the unit cost by 12.3%.

The above only begins to expose the type of information that simulation modeling can reveal. The modeling procedures and certain relations describing the basic processes are well understood. The weak link is data. The geologic--and other--uncertainties with which one must cope provide the real test. Only as more and better data become available will this approach reveal its true utility; hopefully it can be extended to include such things as probabilistic aspects of component failure, statistical reservoir analysis, etc.

End_of_Article - Last_Page 2088------------