On-Site Experimental Study on Low-Temperature Deep Waste Heat Recovery of Actual Flue Gas from the Reformer of Hydrogen Production

Abstract

Improving the energy-saving efficiency of flue gas deep waste heat and reducing the emission of carbon dioxide and other pollutants have been two issues that need to be paid attention to in petrochemical heating furnaces. A hydrogen production reformer with high energy consumption and high carbon emissions in the petroleum refining process affects the thermal and productive efficiency of the hydrogen production, amounts of heat from flue gas are wasted with the exhausted corrosive gas of the reformer, and latent heat is not recovered. To recover the sensible and latent heat from the exhausted gas, a new anti-corrosion, high-efficiency, and low-pressure-drop flue gas condensing heat exchanger (FGCHE) with low consumption and pressure drop was developed. The energy-saving performance was evaluated through on-site measurements and theoretical analysis. The results show that the exhausted gas temperature was reduced from 161.3~175.9 °C to 33.9~38.9 °C after using the new FGCHE to recover waste heat. The energy-saving efficiency and the utilization ratio of flue gas waste heat were 12~16.1% and 74~81.9%, respectively. The latent heat accounted for 41.3~48.1% of the total recovered heat. The exergy efficiency and the total thermal efficiency of the reformer reached 73~86.8% and 95.2~96.6%, respectively. The condensation in the flue gas reduced pollutant emissions (SO₂ and NO_x). This paper provides a practical application reference for the technology development of waste heat recovery and the application of an FGCHE for petrochemical heating furnaces.

1. Introduction

In 2022, China’s total energy consumption was 5.41 billion tons of standard coal, of which industrial energy consumption accounted for about 65% [1]. For industry, around 42% of the total primary energy consumption as waste heat is directly emitted in different forms [2]. Industrial waste heat recovery and hydrogen production are the most promising methods to reduce carbon emissions [3] and realize carbon peaking and carbon neutrality goals. Hydrogen production methods mainly contain hydrocarbon conversion, reforming by-products, and refinery hydrogen-rich gas purification [4,5]. Hydrocarbon conversion to the reformer of hydrogen production plays a main role in hydrogen production, accounting for 80% of total hydrogen production [6]. The reformer determines the amount of hydrogen production and provides raw materials and fuel for the refining process, as shown in Figure 1, while flue gas exhausted from the hydrogen production reformer (HPR) with temperatures of 150~250 °C is exhausted, which is 17~22% of the total heat of fuel consumption. Exhausted waste heat recovery is beneficial to saving energy [7,8], efficiency improvement [9,10], water conservation [11,12], pollution reduction, and carbon reduction [13,14,15].

Heat exchangers (e.g., heat pipe, air preheater, etc.), the most widely used in waste heat recovery technology, significantly affect the amount of heat transfer and thermal efficiency. Therefore, some researchers have studied the heat transfer characteristics of heat exchangers.

Table 1. Application of different heat exchangers for waste heat recovery of flue gas.

Authors	Fuels	Energy-Saving Equipment	Flue Gas Inlet Temperature (°C)	Flue Gas Outlet Temperature (°C)	Inlet Temperature of the Heated Medium (°C)	Outlet Temperature of the Heated Medium (°C)	Waste Heat Recovery Type
Wang et al. [16]	Bituminous coal	Heat pipe heat exchanger	156.8~158.5	129.8~131.2	40.9~42.2	88.5~89.3	Sensible heat recovery
Szulc et al. [17]	Lignite coal	Heat exchanger	155	60	-	90
Wang et al. [18]	Fuel oil	Air preheater	285~347	160~165	80	155~225
Karimi et al. [19]	Natural gas	Economizer	306	203	118	254
Miheli-Bogdani et al. [20]	Natural gas	Air preheater	204	66.93	24	186
Madzivhandila et al. [21]	Syngas	Contact economizer	90	>60	-	-
Shang et al. [25]	Natural gas	Non-contact total heat exchanger	53.5	25.7	0	44.6	Sensible and latent heat recovery
Hong et al. [26]	Natural gas	Tube-type heat exchanger	38~55	28~32	11~13	16~18
Zhang et al. [27]	Natural gas	Direct contact heat exchange	203.5~209.8	45~55	43.7	50.8
Bao et al. [28]	Natural gas	Transport membrane condenser	64.7~93.5	45.2~57.3	19.1–44.6	33.5–56.1

presents the application of different heat exchangers for coal, oil, and natural gas heating furnaces to recover waste heat from flue gas. As shown in Table 1, coal [16,17], fuel oil [18], and gas [19,20,21] heating furnaces have added heat exchangers to recover sensible heat of flue gas, of which the outlet temperature drops to 60~203 °C (higher than the acid dew point temperature). The flue gas contains corrosive gas such as SO_x and NO_x [22,23]. Hence, the low-temperature (lower than the flue gas water dew point temperature) flue gas deep waste heat recovery of refinery heating furnaces is a technical challenge in the petrochemical industry, while for natural gas boilers, the corrosiveness of flue gas is weak. The water dew point temperature of flue gas from natural gas boilers is generally below 57 °C [24]. The exhaust gas temperature of natural gas boilers can be reduced to 25.7~57 °C lower than the water dew point temperature [25,26,27,28], and the sensible and latent heat of flue gas is deeply recovered. However, the latent heat recovery from the HPR of a petroleum refining system has not been reported and there are no projects implemented or on-site measurements. For fuel gas boilers or furnaces, the latent heat of flue gas accounts for more than 10% of the fuel gas input heat [29]. If the latent heat is deeply recovered, the thermal efficiency of the energy-using system can be significantly improved [30]. The flue gas condensate is precipitated in the process of flue gas condensation and recovered after being treated, reducing the mist emission [31]. In addition, the flue gas condensate also absorbs pollutants such as NO_x and SO₂ in the exhausted gas [32], which purifies the flue gas, reduces pollutant emissions, and improves the atmospheric quality. Hence, latent heat recovery from flue gas plays an important role in saving energy, water conservation, and emission reduction.

There are many ways to evaluate the thermal performance of waste heat recovery systems. Energy saving and exergy efficiency are commonly used [33,34]. The energy-saving efficiency of flue gas waste heat is analyzed based on the first law of thermodynamics, while the exergy efficiency is analyzed by combining the first and second laws of thermodynamics. Li et al. [35] conducted an energy and exergy analysis of waste heat recovery from a cogeneration plant and pointed out the optimization direction using thermodynamic analysis. Hernandez et al. [36] adopted exergy to evaluate material efficiency and energy efficiency. Hamedi et al. [37] applied exergy analysis to study the contributions of the olefin plant to total exergy destruction. Utlu et al. [38] employed energy and exergy analysis to obtain the efficiency and heat recovery potential of Turkish industrial sectors. Shariati et al. [39] carried out an exergy analysis on the waste heat recovery section (including the feed preheater part, the pre-reformer reactor, and the reformer reactor) about the sensible heat of flue gas from the steam-natural gas reformer, and the exergy efficiency reached 0.58. However, there are few studies that have investigated the exergy efficiency of the flue gas condensing heat exchanger (FGCHE) used in the HPR.

To fill the above-mentioned gap, the FGCHE is used to recover the sensible and latent waste heat. The temperature of exhausted gas is reduced to below the flue gas water dew point temperature, and then the sensible and latent heat of low-temperature flue gas and the condensed water can be recovered. Meanwhile, the acid-condensed water corrodes the heat exchanger and increases equipment consumption. Low heat-transferred-temperature differences lead to a higher area of the heat exchanger. Additionally, a new anti-corrosion, high-efficiency, and low-pressure-drop FGCHE was developed and applied in the HPR. Meanwhile, the energy-saving performance of the FGCHE was measured on-site. The results of this paper can offer a design reference for similar furnaces or reformers to achieve deep utilization, energy savings, efficiency improvements, and carbon and pollution reduction for low-temperature waste heat recovery. For petroleum refining furnaces with high exhausted gas temperatures, the sensible and latent heat in the flue gas waste heat can be recovered by installing an anti-corrosive FGCHE, saving primary energy and improving the energy utilization ratio of the refining enterprises. Hence, it also provides a feasible solution for low-temperature flue gas deep waste heat to recover the latent heat and condensed water from the flue gas of the petrochemical heating furnaces.

The structure of the following section is as follows. In Section 2, the low-temperature deep waste heat recovery system (LT-DWHRS) for the reformer is designed, and the potential of the low-temperature deep recovery system of waste heat is analyzed; In Section 3, the on-site measurement method and data analysis basis are given; In Section 4, the energy saving performance and exergy efficiency of the FGCHE under different operating conditions are discussed. Additionally, the socio-economic and environmental benefits of the FGCHE are analyzed.

2. System Description

2.1. Low-Temperature Deep Waste Heat Recovery System in a Reformer

The reformer of a 5000 Nm³/h hydrogen production unit was reformed for the LT-DWHRS, as shown in Figure 2. Based on the operating parameters of the reformer and the heat demand of the refinery, the anti-corrosion, high-efficiency, and low-pressure-drop FGCHE was applied at the rear of the air preheater (AP) and draught fan to heat the low-temperature demineralized water, which was used as water replenishment for a self-contained power plant in a petrochemical enterprise.

Pressure swing adsorption (PSA) gas is mainly used for fuel gas combustion in HPRs. The composition and calorific value of the PSA gas are shown in

Table 2. PSA gas components and calorific value.

PSA Gas Components	H2	CH4	CO	CO2	H2S
Volume fraction	29.6%	13.4%	3.4%	53.6%	≤5 mg/m3
Higher heating value (HHV)	9970 kJ/Nm3
Lower heating value (LHV)	8433 kJ/Nm3

. The PSA gas is burned in the HPR to heat the raw material, and produces high-temperature flue gas, which successively passes through the raw material gas preheating section (RMGP), the steam generation section (SGS), the air preheater (AP), the draught fan, the FGCHE, and the stack. The PSA gas consumption, temperature, and components of flue gas after AP are shown in

Table 3. PSA gas consumption, temperature, and components of flue gas from the HPR.

PSA Gas Volume Flow Rate	Temperature	Components
		N2	CO2	H2O	O2	SO2	NOx
Nm3/h	°C	%	%	%	%	mg/Nm3	mg/Nm3
1580~2354	161~191.6	58~62.2	9~9.8	14.5~18.1	3.42~6.23	1.5~2.6	23~38.6

.

2.2. Flue Gas Condensing Heat Exchanger and Its Energy-Saving Potential

The novel anti-corrosion, high-efficiency, and low-pressure-drop FGCHE determines the recovery of flue gas waste heat. The surface of the fin-and-tube for the condensing heat exchanger was modified by anti-corrosion coating, which can prevent corrosion at the dew point of flue gas. The counter current pattern was applied in the condensing heat exchanger, where the flow direction of the flue gas in the FGCHE is from top to bottom, and the cooling water flows from bottom to top. This pattern is conducive to improving the heat exchange efficiency of the heat exchanger. During the heating process, the cooling water releases gases that flow upwards along with the cooling water and are discharged by the exhaust valve at the highest position. The condensed water is released and collected in the tank after the flue gas cools lower than the dew point temperature. Meanwhile, the condensed water flows in the same direction as the flue gas, which drives the flue gas condensate’s timely discharge and improves the heat transfer performance of the FGCHE. The condensed water can be used as supply water after treatment. The exhausted gas pressure of the HPR system is about 100 Pa. The outlet pressure of exhausted gas can overcome the flue gas pressure drop of the FGCHE (less than 30 Pa) without extra power consumption of the draught fan.

Table 4. Energy-saving and carbon-reducing potential analysis of the FGCHE under design conditions.

Contents	Units	Parameters
PSA gas volume flow rate	Nm3/h	2354
Excess air coefficient	/	1.3
Flue gas volume flow rate	Nm3/h	8330
Inlet temperature of flue gas	°C	170
Outlet temperature of flue gas	°C	40
Inlet moisture content of flue gas	g/kg	111.926
Outlet moisture content of flue gas	g/kg	49.532
Cooling water flow rate	m3/h	25
Inlet temperature of cooling water	°C	25
Outlet temperature of cooling water	°C	56
Exhausted heat amount	kW	1203
Latent heat ratio in exhausted heat	%	54.3
Exhausted heat loss (HHV)	%	18.9
Flue gas waste heat recovery amount	kW	902
Utilization ratio of flue gas waste heat	%	64.6
Energy-saving efficiency (HHV)	%	13.8
Dehumidification rate	%	55.7
Flue gas condensed water flow rate	t/d	14.9
Annual operation time	h	8000
Annual saving amount of PSA gas	104 Nm3/a	308.3
Annual reduction in CO2 emissions	t/a	4278

presents the energy-saving and carbon-reducing potential analysis of the FGCHE. Under the design conditions of the FGCHE, the PSA gas volume flow rate, excess air coefficient, and flue gas volume flow rate were 2354 Nm³/h, 1.3, and 8330 Nm³/h, respectively. When the exhausted gas temperature was reduced from 170 °C to 40 °C, the energy saving efficiency, waste heat recovery amount, and utilization ratio of the flue gas were 13.8% (HHV), 902 kW, and 64.4%, respectively. The 25 m³/h low-temperature demineralized water as cooling water was heated from 25 °C to 56 °C, and the flue gas condensed water flow rate was 14.9 t/d. When the annual operation time of the reformer was 8000 h, the annual PSA gas volume flue rate can be saved by 3.083 million Nm³/a, with the CO₂ emissions reduced by 4278 t/a.

3. On-Site Measurement Method

3.1. On-Site Measurement System and Method

The LT-DWHRS was located in Shandong Chambroad Petrochemical Plant. The detection system of the LT-DWHRS comprised PSA gas, flue gas, and cooling water, as shown in Figure 2.

The PSA gas flow rate, temperature, and pressure were measured by the gas flow meter. The flue gas temperatures were detected by T-type armored thermocouples, and the water temperatures were detected by Pt100-type platinum resistances, which were connected to the Agilent. The data cables were used to transmit the data collected by Agilent to the computer. The flue gas flow pressure drop of the FGCHE was read via a digital micromanometer. The flue gas components were detected using a flue gas analyzer. The cooling water flow rate and pressure were detected with ultrasonic flowmeter and pressure gauges, respectively.

Table 5. Instruments and their accuracy.

NO.	Instruments	Companies	Type	Accuracy	Parameters
1	Gas flowmeter	Tancy Instrument Group Co., Ltd. (Wenzhou, China)	FCM-II	±1.5%	PSA gas volume flow rate
2	Ultrasonic flowmeter	Micronics Ltd. (Loudwater, UK)	PF300	±1%	Cooling water volume flow rate
3	Platinum resistance thermometer	Siemens AG (Munich, Germany)	Siemens-Pt100	±0.01 °C	Cooling water temperature
4	Armored thermocouple	Spectris Instrumentation and Systems Shanghai Ltd. (Shanghai, China)	T-type	±0.1 °C	Flue gas temperature
5	Flue gas analyzer	Ecom GmbH (Assamstadt, Germany)	Ecom-J2KN	O2: ±0.2% CO2: ±0.3% NOx: ±2 ppm SO2: ±5 ppm	Flue gas components
6	Digital micromanometer	Ecom GmbH (Assamstadt, Germany)	Ecom-DPH	±3%	Flue gas flow pressure drop
7	Agilent	Agilent Technology Co., Ltd. (Santa Clara, CA, USA)	HP34970A	±0.004%	Data collection
8	Temperature and humidity recorder	Ecom GmbH (Assamstadt, Germany)	Ecom-TFS	±0.1 °C, ±0.1%	Air temperature and humidity
9	Electronic scale	Tianfu Co., Ltd. (Zhengzhou, China)	DT-30K	±1 g	Flue gas condensate flow
10	Stopwatch	Shenzhen Huibo Industry and Trade Co., Ltd. (Shenzhen, China)	PC396	±0.01 s	Time
11	Portable pH Tester	Mettler Toledo (Greifensee, Switzerland)	SevenGo-SG2	±0.01 pH	pH of the condensed water

presents the instruments and the accuracy of the experimental system. Among the testing instruments listed in Table 5, most of the on-site instruments are shown in Figure 2, except for the temperature and humidity recorder, electronic scale, stopwatch, and portable PH tester, which were portable and carried during the on-site tests. The operating parameters of the LT-DWHRS were collected every five minutes, and the detection time was not less than 2 h.

The average temperature of the flue gas was calculated using Equation (1) [40]

$$t_{\mathrm{f}}={\displaystyle \sum }_{i=1}^n\frac{t_{\mathrm{f},i}{A}_i}{N{A}_i}=\frac{{{\displaystyle \sum }}_{i=1}^n{t}_{\mathrm{f},i}}{N}$$

(1)

where t_f and t_f,i are the average and local measuring temperatures of flue gas; A_i is the equivalent area; and N is the measuring number.

3.2. Data Analysis

The fuel gas calorific capacity based on the higher heating value is given by [41]

$$Q_{\mathrm{g}}=H{G}_{\mathrm{g}}$$

(2)

where H and G_g are the higher heating value and volume flow rate of fuel gas, respectively.

Based on the dry flue gas mass flow rate and the flue gas enthalpy difference between the flue gas inlet and standard ambient temperatures, the exhausted heat amount of the HPR is given by

$$Q_{\mathrm{eh}}=m_{\mathrm{f}}(h_{\mathrm{f},\mathrm{in}}-h_{\mathrm{am}})$$

(3)

where h_f,in and m_f are the inlet enthalpy and mass flow rate of dry flue gas; h_am is the flue gas enthalpy value at standard ambient temperature (20 °C).

The exhausted gas heat loss of the HPR is given by [42]

$$q_2=\frac{Q_{\mathrm{eh}}}{Q_{\mathrm{g}}}$$

(4)

where Q_eh is the exhausted heat amount and Q_g is the fuel gas calorific capacity.

Based on the heat balance theory [41], the thermal efficiency of the reformer was calculated as

$${\eta }_1=100-(q_2+q_3+q_4+q_5+q_6)$$

(5)

where q₂, q₃, q₄, q₅, and q₆ are heat loss of exhausted gas, heat loss of unburned fuel gas, heat loss of unburned solid, heat dissipation loss, and other heat loss, respectively.

The total thermal efficiency of the reformer system after adding the FGCHE was calculated by

$${\eta }_{\mathrm{t}}={\eta }_1+\eta$$

(6)

where η is the energy-saving efficiency of the FGCHE.

According to the inversing balance calculation method, the heat amount released by flue gas can be calculated according to flue gas heat release [42]:

$$Q_{\mathrm{f}}=m_{\mathrm{f}}(h_{\mathrm{f},\mathrm{in}}-h_{\mathrm{f},\mathrm{out}})-m_{\mathrm{cw}}{h}_{\mathrm{cw}}$$

(7)

where h_f,out and h_cw are the flue gas outlet enthalpy and flue gas condensed water enthalpy; m_cw is the flue gas condensed water flow rate.

The flue gas condensed water flow rate was calculated as

$$m_{\mathrm{cw}}=m_{\mathrm{f}}\times \frac{(d_{\mathrm{f},\mathrm{in}}-d_{\mathrm{f},\mathrm{out}})}{1000}$$

(8)

where d_f,out and d_f,in are the flue gas outlet and inlet moisture content, respectively.

Releasing heat amount of flue gas includes sensible heat and latent heat. The sensible heat amount of flue gas can be calculated using

$$Q_{\mathrm{f},\mathrm{s}}=c_{\mathrm{p},\mathrm{f}} {\rho }_{\mathrm{f}} V_{\mathrm{f}} G (t_{\mathrm{f},\mathrm{in}}-t_{\mathrm{f},\mathrm{out}})$$

(9)

where c_p,f is the specific heat capacity of flue gas; ρ is the flue gas density under standard conditions; V_f is flue gas volume per unit of fuel gas; and t_f,in and t_f,out are flue gas outlet and inlet temperature, respectively.

The latent heat amount of flue gas can be calculated as

$$Q_{\mathrm{f},\mathrm{l}}=Q_{\mathrm{f}}-Q_{\mathrm{f},\mathrm{s}}$$

(10)

According to the positive balance calculation method, the heat absorption amount of cooling water can be obtained using Equation (11) [40]:

$$Q_{\mathrm{w}}=m_{\mathrm{w}}\left(h_{\mathrm{w},\mathrm{out}}-h_{\mathrm{w},\mathrm{in}}\right)=m_{\mathrm{w}}{c}_{\mathrm{p},\mathrm{w}}\left(t_{\mathrm{w},\mathrm{out}}-t_{\mathrm{w},\mathrm{in}}\right)$$

(11)

where m_w is the mass flow rate of cooling water; h_w,out and h_w,in are the outlet and inlet enthalpies of cooling water; and t_w,out and t_w,in are the outlet and inlet temperatures of cooling water.

The errors between the water-side and flue-gas-side heat balance calculated using Equation (12) were less than 5%.

$$\mathsf{\chi }=\left|\frac{Q_{\mathrm{w}}-Q_{\mathrm{f}}}{Q_{\mathrm{f}}}\right|\times 100\%$$

(12)

Flue gas waste heat recovery amount can be calculated as [43]

$$Q_{\mathrm{h}}=\frac{Q_{\mathrm{w}}+Q_{\mathrm{f}}}{2}$$

(13)

where Q_h is the flue gas waste heat recovery amount.

The utilization ratio of flue gas waste heat of the FGCHE is given by [44]

$${\eta }_{\mathrm{h}}=\frac{Q_{\mathrm{h}}}{Q_{\mathrm{eh}}}$$

(14)

The energy-saving efficiency of the FGCHE is given by [44]

$$\eta =\frac{Q_{\mathrm{h}}}{Q_{\mathrm{g}}}$$

(15)

From the perspective of energy conversion and utilization, the exergy efficiency based on the second law of thermodynamics is often one of the important indicators for evaluating the performance of heat exchangers. The exergy efficiency of the FGCHE was calculated using [45]

$${\eta }_{\mathrm{ex}}=\frac{\Delta E_{\mathrm{w}}}{-\Delta E_{\mathrm{f}}}=-\frac{m_{\mathrm{w}}\left(h_{\mathrm{w},\mathrm{out}}-h_{\mathrm{w},\mathrm{in}}\right)-m_{\mathrm{w}}{T}_0\left(S_{\mathrm{w},\mathrm{out}}-S_{\mathrm{w},\mathrm{in}}\right)}{m_{\mathrm{f}}\left(h_{\mathrm{f},\mathrm{out}}-h_{\mathrm{f},\mathrm{in}}\right)-m_{\mathrm{f}}{T}_0\left(S_{\mathrm{f},\mathrm{out}}-S_{\mathrm{f},\mathrm{in}}\right)}$$

(16)

Herein, ΔE_f and ΔE_w denote the exergy difference of flue gas and cooling water; T₀ is the ambient temperature (293.15 K); S_f,in, S_f,out, and S_w,in, S_w,out are the inlet- and outlet-specific entropy of flue gas and cooling water.

The dew point of flue gas from the HPR was calculated as [46]

$$\lg P_{{\mathrm{H}}_2\mathrm{O}}=a-\frac{b}{t_{\mathrm{d}}+c}$$

(17)

where t_d and P_H2O are the dew point temperature and water vapor partial pressure of flue gas; a, b, and c are the Antoine constants for water vapor pressure.

The dehumidification rate of flue gas at the outlet of the FGCHE was calculated as

$$\psi =\frac{d_{\mathrm{f},\mathrm{in}}-d_{\mathrm{f},\mathrm{out}}}{d_{\mathrm{f},\mathrm{in}}}$$

(18)

Based on the annual PSA gas saving amount and the average concentrations of CO₂, NOx, and SO₂ emissions from the flue gas of the HPR, the CO₂, NO_x, and SO₂ emission reductions of the LT-DWHRS are expressed in Equation (19)–(21), respectively.

$$E_{{\mathrm{CO}}_2}=G_{\mathrm{sg}}·{\epsilon }_{{\mathrm{CO}}_2}$$

(19)

$$E_{{\mathrm{NO}}_{\mathrm{x}}}=G_{\mathrm{sg}}·{\epsilon }_{{\mathrm{NO}}_{\mathrm{x}}}$$

(20)

$$E_{{\mathrm{SO}}_2}=G_{\mathrm{sg}}·{\epsilon }_{{\mathrm{SO}}_2}$$

(21)

where E_CO2, E_NOx, and E_SO2 are the total emissions of CO₂, NO_x, and SO₂ per year depending on variant, respectively; G_sg is the annual saving amount of PSA gas; and ε_CO2, ε_NOx, and ε_SO2 are PSA gas factors for CO₂, NO_x, and SO₂ emissions (ε_CO2 = 1.387, ε_NOx = 1.586 × 10⁻⁴, ε_SO2 = 2.57 × 10⁻⁶), respectively.

The annual saving amount of PSA gas is

$$G_{\mathrm{sg}}=G_{\mathrm{av}}{T}_{\mathrm{ar}}\eta$$

(22)

where G_av is the average volume flow rate of fuel gas and T_ar is the annual running time.

The payback period was evaluated using [47]

$$\mathrm{PBP}=\frac{\mathrm{IC}}{\mathrm{ANCF}}$$

(23)

where PBP, IC, and ANCF are the payback period, the investment cost (including materials and manufacturing costs of the heat exchanger), and the annual net capital flow.

3.3. Uncertainty Analysis

Experimental data measurement were divided into direct measurement and indirect measurement. The former was directly measured with instruments, including the flow rate and temperature of PSA gas and cooling water, condensed water flow rate, and pressure drop of flue gas, etc. The type and accuracy of those instruments are presented in Table 5. Indirect measurement was calculated based on direct measurement values, including exhausted gas waste heat and its recovery amount, utilization ratio, energy saving, and exergy efficiency, etc. According to the uncertainty transitive relations [48] as shown in Equations (24) and (25), the relative uncertainties can be calculated, which depend on both the accuracy of the measuring instruments and testing conditions. The relative uncertainties of Q_eh, q₂, Q_h, η_h, η, η_ex, and ψ are 3.16%, 3.5%, 1.05%, 3.33%, 1.83%, 3.66%, and 1.51%, respectively.

$$\Delta y=\sqrt{{\left(\frac{\partial y}{\partial x_1}\Delta x_1\right)}^2+{\left(\frac{\partial y}{\partial x_2}\Delta x_2\right)}^2+\cdots +{\left(\frac{\partial y}{\partial x_n}\Delta x_n\right)}^2}$$

(24)

$$e=\frac{\Delta y}{y}\times 100\%$$

(25)

4. Results and Discussion

The LT-DWHRS was measured, and the energy-saving performance of the FGCHE under actual operating conditions was obtained.

4.1. Operating Parameters of the LT-DWHRS

As shown in Figure 3a, the PSA gas flow rate of the HPR was 1820~2195 Nm³/h, and the average PSA gas flow rate was 1967.6 Nm³/h. The load rate of the HPR was 77.3~93.2%, fluctuating within a certain range due to the production process changes of the hydrogen production unit. The excess air coefficient was 1.15~1.55 and the average excess air coefficient was 1.37. The flue gas temperatures and flow pressure drop of the FGCHE are shown in Figure 3b,c. After adding the FGCHE at the rear of the reformer, the exhausted gas temperature dropped from 161.3~175.9 °C to 33.9~38.9 °C. The exhausted heat of flue gas heated 22.4 m³/h low-temperature demineralized cooling water from 26~28 °C to 55~60 °C. The difference between the outlet temperature of flue gas and the inlet temperature of cooling water was within 7.7~12.1 °C. The flue gas flow pressure drop of the FGCHE was rarely small with 20~30 Pa. The flue gas flow pressure drop of flue gas fluctuates with the operating load rate of the HPR. The higher load rate of the HPR, the higher amount of flue gas generated by fuel gas combustion, and the higher flue gas flow pressure drop of the FGCHE. The residual exhausted residual pressure behind the draught fan (~100 Pa) can meet the pressure drop requirement of the FGCHE. After the FGCHE cooled the flue gas, the recovering condensed water flow rate of the flue gas was 483.3~680.5 kg/h, equivalent to 11.6~16.3 t/d, as shown in Figure 3d.

4.2. Dew Point and Dehumidification Rate

As shown in Figure 4a, the flue gas dew point temperature was within 51.9~55.9 °C, and the flue gas inlet moisture content of the FGCHE was 97.1~122.2 g/kg with an excess air coefficient at 1.15~1.55. After the exhausted gas temperature reached the dew point temperature, the water vapor in flue gas condensed into condensed water. In this process, it released the latent heat, and the outlet moisture content of the flue gas dropped to 34.7~46.3 g/kg. Figure 4b shows that the condensed water flow rate is proportional to the dehumidification rate of flue gas. The dehumidification rate of flue gas was 55~68.3%. As shown in Figure 4c, with the increase in the excess air coefficient, the dew point and inlet moisture content of flue gas decreased. The inlet moisture content of flue gas decreased, which means the partial pressure of water vapor decreased, whose corresponding saturation temperature decreased. Thus, the flue gas dew point temperature decreased. The dehumidification rate fluctuated with the outlet moisture content fluctuation of the flue gas and showed a decreasing trend with the increasing excess air coefficient. Because the increase in excess air coefficient under the condition that the load rate of the HPR and recovered heat were relatively stable, the volume flow rate of flue gas increased and the flue gas inlet moisture content decreased, while the outlet moisture content of flue gas increased. According to Equation (14), the dehumidification rate of flue gas decreased.

The condensed water of flue gas was collected, and its pH and partial ion concentration were tested. The test results are shown in

Table 6. Test results of flue gas condensed water.

pH	Iron Concentration	Sulfide Concentration	Sulfate Concentration	Chloride Concentration	Nitrate Nitrogen Concentration	Nitrite Nitrogen Concentration
	μg/L	mg/L	mg/L	mg/L	mg/L	mg/L
4.60	39.7	0.005	4.61	1.12	0.35	0.94

. The pH of flue gas condensed water was 4.6. The condensed water absorbed acid gases such as NO_x and SO₂. Thus, the flue gas condensation had a purification effect on NO_x and SO_2, and on other pollutant gases to a certain extent. Considering the acid-condensed water with corrosiveness, the working time of the FGCHE can be prolonged through anti-corrosion surface modification.

4.3. Exhausted Heat and Flue Gas Waste Heat Recovery Amount

4.3.1. Exhausted Heat Amount and Exhausted Heat Loss of the HPR

Figure 5 shows the exhausted heat amount and exhausted heat loss of the HPR. As shown in Figure 5a, the exhausted heat amount of the flue gas was 855~1125 kW, of which the latent heat was 439~595 kW. The total exhausted heat loss of HPR was 15.8~20.2%, and the latent heat accounted for 50.9~54.5% of total exhausted heat. As the excess air coefficient increased, the flue gas volume flow rate increased, while the reformer’s exhausted gas temperature tended to be stable; thus, the exhausted heat amount and exhausted heat loss increased, as shown in Figure 5b.

4.3.2. Recovery Amount and the Utilization Ratio of Flue Gas Waste Heat of the FGCHE

The recovery amount and utilization ratio of flue gas waste heat and its variation with excess air coefficient and outlet temperature of the flue gas are presented in Figure 6. In Figure 6a, the flue gas waste heat recovery amount was 649~859 kW, of which the latent heat was 271~413.3 kW, accounting for 41.3~48.1% of the flue gas waste heat recovery amount. The utilization ratio of flue gas waste heat was 74~81.9%. As shown in Figure 6b, with the excess air coefficient increasing, the recovery amount of flue gas waste heat increased, and the utilization ratio of flue gas waste heat was relatively stable. Because the amount of flue gas increased with the excess air coefficient, the mass flow rate of flue gas increased, increasing flue gas velocity-enhancing heat transfer, and then the recovery amount of flue gas waste heat increased. Meanwhile, the exhausted heat also increased so that the utilization ratio of flue gas waste heat changes were relatively small. In Figure 6c, the recovery amount and utilization ratio of flue gas waste heat decreased with the flue gas outlet temperature. Because the flue gas inlet temperature was relatively stable, the higher the flue gas outlet temperature, the lower the recovery amount and utilization ratio of flue gas waste heat.

4.4. Energy-Saving Efficiency

The comparisons of the on-site and theoretical energy-saving efficiency and thermal efficiency of the reformer are shown in Figure 7a,b. Figure 7a shows that the on-site original thermal efficiency of the HPR was only 80~84% (HHV). The total thermal efficiency of the reformer system after waste heat recovery reached 95.2~96.6% (HHV), including an energy-saving efficiency of 12~16.1% (average 14.4%) (HHV). Because the flow rate of cooling water was relatively stable, the lower the inlet water temperature, the larger the temperature difference between cooling water and flue gas, which led to the enhancement of convection heat transfer. Thus, both the recovery amount and energy-saving efficiency of flue gas waste heat increased. Figure 7b shows the theoretical energy-saving efficiency and thermal efficiency of the HPR at different exhausted gas temperatures under a state of thermal equilibrium. The results show that when the temperature of flue gas dropped from 170 °C to 30~40 °C under the excess air coefficient 1.1~1.4 (flue gas water dew point 53.3~56.6 °C), the theoretical energy-saving efficiency was 13~17% (HHV), the theoretical total thermal efficiency of the HPR was 94.6~97.9% (HHV), and the theoretical utilization ratio of flue gas waste heat was 72~89%. The on-site measurement result of energy-saving efficiency, the utilization ratio of flue gas waste heat, and thermal efficiency of HPR were consistent with the theoretical calculation values. Figure 7c,d presents the variation in energy-saving efficiency with excess air coefficient and outlet temperature of flue gas. The energy-saving efficiency increases with the excess air coefficient, as shown in Figure 7c. Because the flue gas volume flow rate and flue gas waste heat recovery increased with the excess air coefficient, the energy-saving efficiency correspondingly increased. Figure 7d indicates that the energy-saving efficiency decreased along with the raising flue gas outlet temperature. Because the flue gas inlet temperature was relatively stable, the difference between the flue gas inlet and outlet temperature was reduced, so the recovery amount of flue gas waste heat and energy saving rate decreased.

4.5. Exergy Efficiency

Figure 8 presents the exergy efficiency and its variation with the outlet temperature of flue gas. As shown in Figure 8a, the exergy efficiency of the FGCHE reached 73~86.8% under the conditions that the flue gas temperature dropped from 161.3~175.9 °C to 33.9~38.9 °C, and the cooling water temperature rose from 26~28 °C to 55~60 °C. The variations in exergy efficiency and energy-saving efficiency with flue gas outlet temperature are shown in Figure 8b. The exergy efficiency presents an inverse proportional relation to energy-saving efficiency. As flue gas outlet temperature decreased, the energy-saving efficiency increased, while exergy efficiency slightly decreased.

4.6. Analysis of Socio-Economic and Environmental Benefits

An LT-DWHRS was used to heat demineralized cooling water for the captive power plant at a low temperature of 26~28 °C. The exhausted gas temperature was reduced from 161.3~175.9 °C to 33.9~38.9 °C, and the energy-saving efficiency was 12~16.1% (HHV). The results of energy-saving, socio-economic, and environmental benefits of the FGCHE for an HPR system are shown in Figure 9.

From the perspective of energy saving benefits, the average consumption of PSA gas for the reformer combustion of the hydrogen production unit reached 1967.6 Nm³/h, and the energy-saving efficiency was up to 12~16.1% (average 14.4%). The unit price of PSA gas was 1.6 RMB/Nm³. When the annual operation time was up to 8000 h, the annual PSA gas consumption of the HPR was saved by 1.889~2.534 million Nm³/a, equivalent to 643.7~863.6 tce/a. The annual PSA gas saving fee reached 3.022~4.055 million RMB/a. The investment cost of the waste heat recovery system was 1.8 million RMB. The payback period was about 5.3~7.2 months (average 5.9 months). The energy saving and economic benefits are remarkable.

From the perspective of environmental performance, the annual savings for PSA gas reduced carbon dioxide emissions by 2620~3516 t/a. The annual pollutant emissions of NO_x and SO₂ were reduced by 299.6~401.9 kg/a and 17.9~24.0 kg/a. Thus, the LT-DWHRS can effectively reduce carbon dioxide, pollutant emissions, and improve the air quality of the refinery area.

5. Summary and Conclusions

Considering the contradiction between energy growth and the achievement of carbon peaking and carbon neutrality, the deep waste heat recovery of industrial exhaust gas, with the aim of efficiency improvement and heat loss reduction like that of the LT-DWHRS applied in this study, is a promising and practical exploration for sustainable development. In this regard, a new FGCHE is proposed to deeply recover the low-temperature waste heat of flue gas from the HPR. The on-site test and practical application of FGCHE are investigated to evaluate its waste heat recovery performance. The socio-economic and environmental benefits are also analyzed. The key conclusions are drawn as follows:

(1)

The FGCHE reduced the exhausted gas temperature from 161.3~175.9 °C to 33.9~38.9 °C, and the utilization ratio of flue gas waste heat was 74~81.9%, of which the latent heat accounted for 41.3~48.1%. The energy-saving efficiency was 12~16.1% (HHV), and the total thermal efficiency of the HPR system reached 95.2~96.6% (HHV). The exergy efficiency was 73~86.8%. The flue gas pressure drop of the FGCHE was rarely 20~30 Pa.

(2)

After the exhausted gas temperature drops to the dew point, 11.6~16.3 t/d flue gas condensed water can be recovered. The condensed water can purify NO_x and SO₂ in the flue gas to a certain extent.

(3)

The annual PSA gas saving was 1.889~2.534 million Nm³/a, which is equivalent to 643.7~863.6 tce/a. Around 3.022~4.055 million RMB/a of the fuel gas cost can be saved after using the FGCHE. The CO₂, NO_x, and SO₂ emissions were reduced by 2620~3516 t/a, 299.6~401.9 kg/a, and 17.9~24.0 kg/a, respectively. The payback period was 5.3~7.2 months. The energy saving, carbon reduction, and economic benefits of the FGCHE for reformers are significant.

China’s crude oil processing volume reached 703.554 million tons in 2021, equivalent to 100.597 million tons of standard coal. If the energy consumption of refining accounts for 8%~10% of total crude oil processing volume, then the refining energy consumption reaches 80.407~100.51 million tons of standard coal. If the flue gas deep waste heat recovery technology is extended to the whole refinery heating furnaces, and the thermal efficiency of the heating furnaces are improved by 12%, then the annual saving amount of standard coal will reach 964.9~12.06 million tons/a, which has huge energy-saving, socio-economic, and environmental benefits.