1. Introduction
Heat recovery technologies have attracted increasing attention around the world due to the increase in fuel prices, rigorous emission regulations, and concerns about environmental problems associated with the use of existing energy systems [
1,
2]. A number of waste heat recovery technologies have been proposed, such as organic Rankine cycles (ORCs) [
3,
4], Rankine cycles [
5], absorption/adsorption cycles [
6], Kalina cycles [
7], thermoacoustic engines [
8], and thermofluidic oscillators [
9,
10]. Among the stated cycles, ORC-based technologies are one of the most promising solutions because of their relatively high efficiency and simple configurations [
11,
12]. ORC technologies have been widely investigated and can be used to recover heat sources such as the waste heat from internal combustion engines [
13,
14], solar energy [
15,
16], geothermal, and other industrial applications [
3,
17,
18]. However, the heat transfer process of a conventional ORC system in the evaporator suffers from poor performance of temperature matching between the heat source and working fluid due to the phase-change evaporating process, causing relatively high exergy destruction and poor thermodynamic performance [
19,
20]. Using mixture fluids can potentially solve the stated problem due to the non-isothermal process in the evaporator [
21,
22]. However, it is difficult to develop the mixture fluids, and their thermodynamic performance need to be further investigated [
23].
The trilateral cycle (TLC) was previously proposed, which has a better thermal match from the heat source to the working fluid than conventional ORCs [
7,
24,
25]. The characteristic of the TLC is that the working fluid directly flows into the expander after being heated in the evaporator without any evaporating process. Therefore, the temperature difference between the working fluid and heat source in the evaporator can be smaller than that of a conventional ORC, leading to an overall smaller exergy loss in the evaporator [
26]. Johann Fischer compared the performance of a TLC and an ORC for waste heat recovery [
27]. Results showed the exergy efficiency of the TLC can be 14–29% higher than that of a conventional ORC under the same operational conditions. M. Yari et al. studied TLC, ORC, and Kalina cycle using an exergoeconomic analysis method [
7]. Results indicated TLC has better power output performance than that of ORC and Kalina cycle, but that the device cost is mainly determined by the isentropic efficiency of the selected expansion machine. A comparison of different pure working fluids illustrated that adopting n-butane can reduce the cost of the TLC system. Ramon Ferreiro Garcia et al. [
28] pointed out that within a range of relatively low operating temperatures, the TLC can obtain an overall thermal efficiency at 44.9% compared to 33.9% of a Carnot cycle under the same high and low-temperature heat reservoir conditions [
28]. Cipollone et al. [
29] conducted a thermodynamic analysis of the TLC system considering the potential to use rotary positive displacement expanders as the expansion device to achieve two-phase expansion of the TLC. The reported feasibility study mainly focusing on the effects of using pure and mixture fluids on volumetric expansion machines under 10 kW power output [
29]. The importance of considering the built-in volume expansion ratio for the TLC system was addressed, and results suggested that a relatively large built-in volume expansion ratio is required to produce a great specific work [
29], which means that in order to achieve overall good performance, volumetric expansion machines are generally not recommend for use with TLC systems.
Previous studies indicated the overall performance of the heat recovery system using the TLC system can be higher than that of a conventional ORC system under the same operational conditions. However, previously-reported research on the TLC is mainly focusing on the performance studies of a single loop system. A single-loop TLC system will require a large evaporator heat exchanger to effectively and efficiently recover high-temperature heat sources such as from industrial processes and the transport sector. Moreover, a single-loop TLC cannot be used for multi-heat source recovery, which has less flexibility compared to a dual-loop cycle or cascaded systems. Extensive research efforts have been made on the dual-loop ORC system, because the system can effectively and efficiently convert dual heat sources into useful power. Moreover, dual-loop ORC technology has the advantage of being able to comprehensively recover a single heat source with a high-temperature profile. Wang et al. [
30] first proposed the concept of a dual-loop ORC and investigated the performance of the system recovering engine exhaust and coolant energy from a gasoline engine. The results showed that the absolute effective thermal efficiency for the objective engine was improved by 3~6% under all operating conditions. Zhang et al. [
31] designed a novel dual-loop ORC to recover the waste heat of exhaust gas, intake air, and coolant energy from a light-duty diesel engine. Results indicated that the effective power of the diesel engine can be improved by 14~16% and 38~43%, corresponding to peak effective thermal efficiency and low load (and high speed) conditions. Other representative works have been conducted by Shu and Tian et al. [
22,
32,
33] and Song and Gu et al. [
34,
35]. Results showed that the system performed better at a high operating load, and R1234yf was found to be the best working fluid. Using water as the working fluid of the HT loop, the system obtained maximum net output power (39.67 kW) and exergy efficiency (42.98%) [
32,
33]. Song et al. [
35] conducted a parametric study based on a dual-loop ORC system. Key parameters including condensation temperature of the high-temperature cycle and working fluids for high-temperature and low-temperature cycles were considered. The results based on a simulation showed that the maximum net power output could achieve as high as 111.2 kW by the proposed dual-loop ORC, increasing the engine power by 11.2% with cyclohexane and R236fa as high-temperature and low-temperature working fluid, respectively [
35].
In summary, previous studies have compared the TLC with other power generation cycles, and have proven the TLC has great potential for heat recovery. Meanwhile, the superiority of using dual-loop ORC has been revealed, which can be used to effectively recover multi-heat sources, although large exergy loss still exists in the evaporator of the dual-loop ORC system. Therefore, in this study, the TLC has been introduced into a conventional dual-loop ORC system to form an innovative cascade system with the purpose of adding flexibility of the TLC for the potential application of multi-heat source recovery, and improving the overall energy efficiency of a conventional dual-loop ORC system. This study aims to study the thermodynamic performance of the proposed TLC-ORC system and conduct a comparative investigation of the conventional dual-loop ORC system under the same operational conditions. The key factors of the systems investigated in this work include the net power output, thermal efficiency, exergy efficiency, HT working fluids, LT working fluids, and the evaporating temperature of HT and LT cycles.
2. System Description
Figure 1 shows the layout of the proposed combined TLC-ORC waste heat recovery system, which consists of a High-Temperature (HT) loop and a Low-Temperature (LT) loop. The schematic diagram of the TLC-ORC is the same as the conventional dual-loop ORC system illustrated in
Figure 1. The TLC is used to replace the HT loop, as shown in the dashed black line. The condensation heat from the TLC is used to preheat the working fluid in the LT loop. In order to understand and compare the working process of the TLC-ORC and a conventional dual-loop ORC system, the
T-s diagrams of two systems are respectively presented in
Figure 2 and
Figure 3. The other technical challenge that should be noted is the expansion machine to be used in the TLC system, which would be forced to operate under two-phase conditions. Recently, works have confirmed the feasibility of using a screw-type expander [
27,
36] or a reciprocating engine [
37] for the TLC system.
A comparison of the working principle between the proposed TLC-ORC and a conventional dual-loop ORC is illustrated in
Figure 2 and
Figure 3. The HT cycle of the TLC-ORC system maintains the working fluid as the liquid state before entering the expansion machine, which can potentially achieve a better thermal match of the heat source conditions. The detailed working process of the TLC is explained as follows, and is shown in
Figure 2a. In state 1, working fluid is in a saturated liquid state at temperature
T1 and vapor pressure
p1. Then, the pressure is increased to
p2 by pump 1 at the state 2 in the homogeneous liquid. The liquid-phase working fluid enters the heat exchanger where it is heated to boiling point at pressure
p2, which is state 3. The temperature
T3 is the boiling temperature at pressure
p2. Starting from state 3, the working fluid directly enters the two-phase expander. In the two-phase expander, the working liquid expands into the wet vapor region, and gradually reaches the wet saturated vapor state (state 4), as shown in
Figure 2a. The authors would like to point out the dryness of the working fluid at state 4 is determined by the calculation after the expansion process and condensation pressure controlled by the condenser. The working fluid at state 4 can be either saturated vapor, the wet state, or within the two-phase region. The wet saturated vapor is condensed until it reaches state 1. For the LT cycle in the TLC-ORC system, LT working fluid starts at state 5 with saturated liquid. Then, it is pressured to state 2 with pressure
p6 by pump 2. The LT working fluid absorbs residual heat from HT working fluid in condenser 1, and continues to absorb the waste heat of exhaust in evaporator 2. At the outlet of evaporator 2 (state 8), LT working fluid is saturated vapor; then, it flows into expander 2 to convert thermal energy to effective work by expanding the process. At state 9, LT working fluid is low pressure and it is condensed to saturated liquid at state 5 by cooling water. The working process of a traditional dual-loop ORC is similar to that of the proposed TLC-ORC. The main differences are the evaporating and condensing processes of the HT working fluid. The detailed working process of conventional dual-loop ORC can be found in reference [
38].