6.2.1. Scheme 1: Deaerator is Involved in the System

In this scheme, steam is only used at stream ends, while the deaerator is employed as cited in [18], maintaining its operating temperature at 158 ◦C. Figure 2 presents the structure result of Scenario 1, which features the TAC of 3,599,059 \$·y<sup>−</sup>1.As shown, cold process stream C1 is heated by hot process streams H1, H2 and H3, C2 is heated by H2 and H4, three heaters are needed at stream ends of C2, C3 and C4. In total, 2550 kW LPS, 6300 kW MPS and 3228 kW HPS are generated from utility system and sent to HEN. Besides, exhaust steam is extracted and condensed at turbine end, mixing with condensates in deaerator and maintaining its operating temperature.

**Figure 2.** Structure with minimum TAC of Scenario 1.

Figure 3 shows the solution of Scenario 2, discarding the location limit on steam heaters. The TAC of the solution is 2,866,637 \$·y<sup>−</sup>1.

**Figure 3.** Structure with minimum TAC of Scenario 2.

As shown, C1 is heated by H1 and H3, C2 is heated by H2 and H4, C3 is heated by H3. It is worth noting that the originally required higher pressure steam in Figure 2 is replaced by lower pressure steam with the load of 12,078 kW of LPS. One heater of the two is located within the inter-stage of C1. This illustrates that the appropriate use of steam considering the inner- and inter-stage location is able to enhance the use of lower pressure steam.

The cost composition of the two scenarios is summarized in Table 2. As indicated, fuel cost of Scenario 2 (3,240,049 \$·y<sup>−</sup>1) is 32.2% lower than that of Scenario 1 (4,782,431 \$·y<sup>−</sup>1). This is because only higher pressure steam can be used at stream ends in Scenario 1 due to the temperature constraints, so the resultant high temperature condensate needs additional extracted and condensed exhaust steam to maintain the operating temperature of deaerator, leaning to the greater fuel consumption. Accordingly, the boiler cost (212,203 \$·y−1) and turbine cost (152,718 \$·y−1) of Scenario 1 are both higher than Scenario 1, while for heat exchanger investment, higher pressure steam is replaced by low pressure steam, increasing the heat transfer temperature difference and increasing heat exchanger area,

so by adding all concerned components, equipment investment of the two scenarios make nearly no difference. In spite of the decreased power generation profit in Scenario 2, fuel costs account for most of the investment, thus finally, TAC of Scenario 2 presents a 20.4% lower than Scenario 1. Conclusion can be made that the improved HEN superstructure provides more possibilities for the selection of multiple utilities and lower pressure steam is more economical in this system.


6.2.2. Scheme 2: Deaerator is not Involved in the System

In Scheme 2, condensates of the steam in different levels are recovered without flowing through the deaerator. Superheated steam is desuperheated by the corresponding condensates. Two scenarios with different HEN superstructures (Scenario 3 and Scenario 4) are studied.

Figure <sup>4</sup> presents the obtained configuration of Scenario 3, with TAC of 2,774,184 \$·y<sup>−</sup>1. As indicated, HPS, 3228 kW, MPS, 6300 kW and LPS, 2550 kW are used at stream ends of C2, C3 and C4 respectively. Figure 5 is the configuration solution of Scenario 4, which costs 2,670,969 \$ per year. As shown, totally 10,848 kW of LPS and 1230 kW of MPS are extracted from utility system. Besides being used within inner-stage of C1, LPS is also allocated sequentially after MPS on C2 within the inter-stage. That is to say, partial MPS is replaced by LPS in this scenario. From the resulted configuration it is implies that lower pressure steam are located within inner- and inter-stage in the improved HEN superstructure.

**Figure 4.** Structure with minimum TAC of Scenario 3.

Table <sup>3</sup> summarized the cost composition of Scenario 3 and 4. Fuel cost of Scenario 4 (2,852,680 \$·y<sup>−</sup>1) is increased by 5.5% compared to Scenario 3 (2,704,636 \$·y−1). As for the equipment investment, utilization of lower pressure steam in Scenario 4 results in a 10.6% higher heat exchanger cost (690,124 \$·y−1) than that of Scenario 3 (624,017 \$·y−1). Boiler and turbine investments of the two scenarios are nearly the same. However, more power is generated when producing lower pressure steam through the turbine in Scenario 3, due to which the profit of power generation in Scenario

4 (1,156,928 \$·y−1) is 39.3% higher. Combining all the above costs, the improved HEN structure gains lower TAC in total, about 3.7% decrease. It can also be concluded that lower pressure steam is more economical.

**Figure 5.** Structure with minimum TAC of Scenario 4.


The two schemes demonstrate that the generation and utilization of utilities have great effects on the HEN structure and equipment investment. In the design where the deaerator is employed in the utility system as cited, more exhaust steam needs to be extracted and condensed to guarantee the operating temperature by mixing with condensed water when higher pressure steam is extracted from the turbine. This will result in more fuel consumption and higher TAC because of the expensive fuel expense. Thus, using lower pressure steam is more economical. In the structure where condensed water is recovered without a deaerator, more power is generated and less heat is supplied when producing lower pressure steam with the same fuel consumption compared with higher one. It is proved that lower pressure is more economical when supplying the same amount of heat to the HEN, although with lower temperature grade. From the above analysis, it is demonstrated that, on the one hand, allowing heaters within inner- and inter-stages is able to generate a better management for utility utilization than the case of only having heaters at stream ends, because steam selections are less restricted by temperature difference constraints and more LPS can be used, which is conducive to the use of energy cascade utilization. On the other hand, a better trade-off among equipment investment, power generation and fuel consumption can be made through the optimization of the whole system with the TAC target.
