From this point in the study, the prospects for finding a revamp which avoidedusing the extra fired heater appeared very good indeed and provided a tremendoustimulus to the operating company’s and the contractor’s engineers to find such adesign.The message is reinforced by looking at the variation of both energy use and heatexchanger area with ∆Tmin (Figure 9.4 and Figure 9.5 respectively). Both the conractor’s network and the increased-area variant are clearly a long way above the l energy consumption for the plant at a ∆Tmin of 7°C, and they are also significantly above the area targets, which suggests that the existing heat exchangers arenot deployed in the most effective way. Again, this gives strong encouragement thata more cost-effective design can be found.The question that may be puzzling us is, how did an experienced contractor apparently get it wrong? And why does the network layout of Figure 9.2 give such poorresults? There seems nothing obviously wrong with it, and the streams have beenmatched roughly in descending order of temperature, which should logically give thebest temperature driving forces and best heat recovery. However, when we plot thecomposite curves (Figure 9.6) and the grand composite curve (GCC) (Figure 9.7) webegin to get some clues. The pinch, at a shifted temperature of 173°C, is not sharp;there is a very long region of constrained temperature driving forces on either side, andthis means that any non-optimal match in this region is very likely to transfer heatacross the pinch or severely squeeze the driving forces in other neighbouring matches.