One of the major problems in geotechnical engineering and particularly in tunneling is the planning of exploration, the selection of methods and locations of exploration. The goal of exploration is to reduce our uncertainty about the geologic conditions, but when planning the exploration these uncertainties still exist. An additional consideration is the meaning of "reduction of uncertainty". We have to be able to put a value on this reduction of uncertainty and this can be done only by determining the effect of reduced uncertainty on construction cost. Exploration planning is thus a classic problem of decision under uncertainly where due to the uncertainty of the environment, several actions are possible and where decisions have to be based on potential cost consequences of these actions. The application of decision analysis to the exploration problem is thus strongly indicated. It is, however, fair to say that only in mineral exploration--where the problem is somewhat more straightforward has the decision analysis approach been used. In geotechnical engineering and tunneling in particular, we are only at the very beginning. One of the reasons for this lagging behind is the practical difficulty in geotechnical problems of locating critical areas which should be explored. Approaches that have been used up to now are either based on intuition or on simplifications (Lindner et al, 1975) and contain inconsistencies that may affect the decision. The present paper will introduce a consistent use of decision analysis in exploration for rock tunneling. A brief overview of the principles of the decision analysis approach will be given first, followed by specific application to tunnel exploration which includes details on the analysis. An example will conclude the paper.
Two main topics will be covered in this section: 1) the principle of decision analysis and 2) some of the major elements of decision theory.
Figure 1 is a presentation of the Decision Analysis cycle consisting of the deterministic, probabilistic and information phase leading to the decision, which can be a decision to collect more information thus leading to a cyclic process. Each of the major phases consists of the steps also shown in Figure 1. Most steps will be described in detail in the application of Decision Analysis to tunnel exploration (Section 3) and the only comments made here are to clarify the overall concept.
Figure 1 - The Decision Analysis Cycle (after Stoël van Holstein, 1973) (available in full paper)
In the e one actually goes through the possible courses of action and quantitatively determines their outcome (1-3). In each course of action there are variables that influence the outcome: decision variables that can be controlled by the decision maker and state variables representing the uncontrollable environment (4). The variables and the outcomes have to be related to each other in a model (5). The outcomes have then to be compared with each other which usually involves cost criteria and/or time criteria (6,7). The sensitivity study (9) will make it possible to identify the most important variables. The e is basically a "revision" of the deterministic phase introducing now the uncertainty of the variables.