E = k_{0} + k_{1}.D -----------(1)
where D is the measured value of the driving factor (tonnes produced, or degree days, etc.) and k_{0} and k_{1} are constants unique to the process being monitored. If there are two or more driving factors D_{1}, D_{2}, D_{3}, and so on, the expression for calculating expected consumption becomes
E = k_{0} + k_{1}.D_{1} + k_{2}.D_{2} + k_{3}.D_{3} -----------(2) See numerical example
Your M&T spreadsheet or other software should be able to handle this calculation perfectly well, once appropriate values of the constant k_{0}, k_{1}, k_{2}, and so on have been found. Multiple regression analysis is one method of establishing these constants, and is a facility found in spreadsheet programs and the better kinds of dedicated M&T software.
Such a multi-driver target cannot be expressed diagramatically but it is possible to manipulate it to make it look like a single-factor target (which can then be shown as a two-dimensional scatter diagram). To do this one must decide which is the 'dominant' driving factor, and subtract from expected consumption the estimated effects of the remaining factors. For example, referring to equation (2), driving factor 2 can be assumed to account for k_{2}.D_{2} units of consumption while driving factor 3 accounts for k_{3}.D_{3} units. This expression...
(E - k_{2}.D_{2} - k_{3}.D_{3})
...represents expected consumption net of the effects of factors 2 and 3. For convenience we can call the expression E'. Rearranging equation 2 gives us...
(E - k_{2}.D_{2} - k_{3}.D_{3}) = k_{0} + k_{1}.D_{1} , or
E' = k_{0} + k_{1}.D_{1} -----------(3)
Equation 3 has the same form as equation 1 and represents a straight line.