Confirmation and prevention of vapor bypass in absorption heat pump with U-pipe pressure separation device caused by upward side two-phase flow

Highlights

• Indication that two-phase flow in U-pipe causes vapor bypass in absorption heat pump.

• Observation of vapor bypass in U-Pipe through experiment under vacuum condition.

• Evaluation of effects of vapor bypass in AHP through field tests.

• New design method to avoid vapor bypass and improve performance by over 6 %.

Abstract

A U-pipe is the key pressure separation device in a lithium bromide absorption heat pump (AHP) connecting the high-pressure condenser and the low-pressure evaporator. A recent study reported that two-phase flow occurs in the upward side of the U-pipe, which reduces the overall density of the fluid and impairs the pressure separation ability of the U-pipe. The liquid seal function of the U-pipe fails, and the vapor can flow directly from the high-pressure condenser to the low-pressure evaporator, causing a vapor bypass problem. This study experimentally investigates the U-pipe vapor bypass mass flow rate under vacuum conditions. The mass flow rate ratio of the bypassed vapor to the main liquid fluid in the U-Pipe can be as high as 10 %. The effect of this bypassed vapor on the AHP performance is investigated through simulations and field tests. The cooling coefficient of performance (COP) drops by 0.05 to 0.1 when vapor bypass occurs, providing evidence for the existence of U-pipe vapor bypass in AHPs. A practical method is suggested to solve this problem and to improve the cooling COP of the AHP by more than 6%.

Introduction

The lithium bromide absorption heat pump (AHP) is an environmentally friendly thermal-driven heat pump that is widely used for waste heat recovery purposes. In China, AHPs have been used in the substations of district heating (DH) systems in recent years to reduce return primary water temperature and to collect more low-grade waste heat to realize clean-energy heating (Xie and Jiang, 2017). An AHP works with a plate heat exchanger to finish the heat exchange from the hot primary water to the secondary water in a heating substation (Zhu et al., 2016; Wang et al., 2018). The hot primary water from the power plant flows through the generator, the plate heat exchanger, and the evaporator in succession to release heat. Then it returns to the power plants. The secondary water from the buildings flows through the absorber, the condenser, and the plate heat exchanger to absorb heat. Then it flows into buildings for space heating. This novel application boosts the demand of AHPs. Since the market demand for AHP cooling capacity is more than 5000 MW annually, even a small improvement in AHP performance will make a significant impact on the energy savings and emission reductions of heating.

In AHPs and absorption chillers, a pressure separation device is used between the condenser and the evaporator. The device allows for a liquid refrigerant flow channel but blocks the vapor flow channel to maintain a vapor pressure difference between the condenser and the evaporator. This device is essential for keeping a stable chamber pressure and refrigerant fluid flow rate.

The pressure drop of a fluid flowing through a pressure separation device can be divided into three parts (Zhu et al., 2020): the friction pressure drop, the acceleration pressure drop, and the gravitational pressure drop. The first two parts are related to the velocity, and the last part is related to a height change. Based on this principle, two types of pressure separation devices have been designed. The first type, including orifice plates and expansion valves, relies on velocity-related pressure drop through the device to create a pressure difference (Manu et al., 2018; Somers et al., 2011). A flow coefficient is defined to describe the relationship between the pressure drop and the quadratic power of the flow velocity through the orifice or the valve. The second type of pressure separation device is the U-pipe, as shown in Fig. 1. A U-pipe relies on the height difference in liquid level between the upward and downward sides of the U-pipe to create a gravitational pressure drop (Zhu et al., 2020; Shin et al., 2009).

These two types of devices have different features in off-design conditions. The pressure difference and liquid flow rate changes are highly related to each other for conventional orifice plates. The control range of flow rate and pressure difference is limited. The expansion valve has a similar pressure separation principle as the orifice plate but provides a wider control range since the flow coefficient is changeable. However, the control of flow coefficient requires additional control strategy and cost. For the U-pipe, the velocity-related pressure drop accounts for a tiny proportion of the total pressure drop. Therefore, one can change either the pressure difference or the flow rate of the U-pipe and keep the other one stable in off-design conditions. Compared with the orifice plate, the U-pipe has a broader control range. Compared with the expansion valve, the U-pipe requires no control strategy and very low cost.

For a traditional absorption chiller, the pressure difference between the condenser and the evaporator varies in a small range in off-design conditions, as shown in Table 1. These conditions work well with the first type of pressure separation device (e.g., orifice plates and expansion valves), which is widely used in compact designs. However, AHPs applied in district heating (DH) systems should work continuously through the whole heating season. The heat rate and temperature level changes widely according to the building heating load and the heating company's control strategy. In this case, the pressure difference between condenser and evaporator varies in a much wider range; therefore, the first type device is no longer suitable for supplying pressure differences and maintaining a stable flow rate. For this reason, the U-pipe is chosen as a better alternative for use in current AHP designs (Yi et al., 2020; Zhu et al., 2016; Vinther et al., 2015; Shin et al., 2009).

The previous understanding of U-pipe flow in an AHP is shown in Fig. 1. The downward inlet side connects to the high-pressure chamber, while the upward outlet side connects to the low-pressure chamber. The downward side of the U-pipe consists of a “free-falling zone” and a liquid zone. The inlet fluid flows as falling film on the pipe wall inner surface in the free-falling zone, and then hits a free liquid surface between the two zones and enters the liquid zone. The upward side of U-pipe is filled with fluid. Researchers and engineers previously believed that the condensing water inside the U-pipe was in a liquid state. The required length of U-pipe is equal to the designed pressure difference, $P_1-P_2$, divided by the product of the liquid density, $\rho$, and gravitational acceleration,$g$ (Shin et al., 2009), as shown in Fig. 1. However, a recent experimental observation of the U-pipe flow inside an AHP under vacuum conditions revealed that flash evaporation occurs at the upward side of the U-pipe (Zhu et al., 2020). The fluid flow in the upward side rapidly changes from single liquid flow to two-phase flow, as shown in Fig. 2a.

When two-phase flow forms in the upward side of the U-pipe, the overall density of the fluid decreases. In vacuum conditions, the difference in properties between a vapor and liquid is huge. The void fraction can be very high even when the vapor mass fraction is low, and the overall density of the vapor can be as small as one-half to one-fifth of the liquid density. This reduction in the overall density impairs the pressure separation capability of the designed U-pipe. For example, a 1.5-meter-tall U-pipe can provide a pressure separation of approximately 15 kPa if the fluid is pure liquid. The same U-pipe can only provide a pressure separation of 3.0 to 7.5 kPa if the fluid flow is a two-phase flow. When the pressure difference between the condenser and evaporator is higher, the liquid zone on the downward side of the U-pipe disappears. The U-pipe is thus unable to separate the vapor of the two chambers, and annular flow or dispersed flow forms in the U-pipe, as shown in Fig. 2b. When this occurs, vapor flows directly from the condenser to evaporator with a very high speed. We call this the vapor bypass condition.

When there is vapor flowing from the condenser to the evaporator, the vapor does not participate in the condensation–evaporation process and is thus wasted. The cooling rate in the evaporator decreases, and the cooling coefficient of performance (COP) of the AHP is reduced.

The cooling COP is one of the most important and direct indices to evaluate the performance of AHPs collecting waste heat, and many studies have focused on the factors influencing the COP. For thermodynamic studies, Herold (1999) provided a zero-order model that predicts COP using heat input and output temperature level. Zhu et al. (2019) provided a unified model for both absorption and the adsorption cycle. The COP is highly dependent on the heat exchange effectiveness of the solution heat exchanger. Xie and Jiang (2015) proposed a new thermodynamic model of an AHP based on internal solution circulation. They pointed out that the solution mixing heat is another factor that influences the COP. More detailed investigations of factors influencing COP have also been conducted by numerical simulation and experimental testing. Both Manu et al. (2018) and Kühn et al. (2005) simulated and tested the effect of the solution circulation ratio on COP. Higher circulation ratio increases sensible heat exchange of the solution and thus reduces COP. Bakhtiari et al. (2011) indicated that the hot driven fluid temperature has little effect on COP, based on experiment and simulation results. However, the cooling fluid temperature and flow rate have a great effect on COP.

The effect of vapor bypass from condenser to evaporator on COP is rarely mentioned in previous studies. This is not surprising because the U-pipe is supposed to be well designed to prevent vapor bypass. In fact, AHP manufacturers all have their own minimum height requirement for U-pipes in AHP design (for example, 1.5 m to 2 m, depending on the working condition) to prevent the vapor bypass problem. However, engineers fail to take the two-phase flow characteristic of the upward side of the U-pipe into consideration. The actual U-pipe height requirement would be higher, but this requirement can hardly be realized since it is close to, or larger than, the total height of most AHPs. Therefore, U-pipe vapor bypass may happen in AHPs under middle and high load conditions.

This paper will focus on the U-pipe vapor bypass problem as a result of the newly discovered two-phase flow characteristic inside the upward side of a U-pipe under vacuum conditions. In Part 2, an experiment was conducted to observe the vapor bypass phenomenon inside a U-pipe and to investigate the vapor bypass flow rate under the AHP working condition range. In Part 3, a field test of an on-site AHP applied in DH system was completed to prove the existence of U-pipe vapor bypass in actural AHPs. A possible method is suggested at the end of the paper to prevent the vapor bypass problem.