Drying plays an integral role in almost every industrial branch and consumes vast amounts of energy. In the industrialised regions, drying processes currently account for 10-25% of the total industrial energy consumption. Drying is responsible for up to 70% of the total energy consumption in the wood industry. This percentage is 30% for the paper processing industry and 50% for the textile industry. Furthermore, most drying installations rely on fossil fuels to produce heat, contributing significant amounts of greenhouse gas emissions.
It is well established that heat pump dryers provide significantly higher energy efficiencies and lower greenhouse gas emissions than conventional dryers. The hot exhaust air from dryers contains large amounts of water vapour which is a source of latent heat. Heat pump (HP) drying provides a way to recover large amounts of this heat, reducing energy consumption by 60-80% at a given temperature. Heat pumps have been around for several decades, yet research and development of heat pump drying technology has only recently advanced to industrial applications.
To explore the application of this technology in industrial drying processes, researchers have embarked on the HP4Drying project in collaboration with partners such as Fraunhofer Institute for Wood Research, Wilhelm-Klauditz-Institut WKI. This project aims to harness the energetic and environmental benefits of heat pumps by integrating them into industrial drying processes. To this end, a heat-pump-assisted dryer was developed.
This paper describes the development and simulation of an electrically driven heat pump dryer test bench, in which the drying process is replicated by means of water evaporation.
Test Bench Description
The test bench aims to demonstrate various optimisation possibilities and expand the field of application towards a higher temperature range. The air cycle optimisations include an adjustable recirculation ratio and air-to-air heat exchanger between input and output. When an electrically driven heat pump with a thermal capacity of 10kW was integrated, cycle optimisations include the use of a suction line heat exchanger, subcooler, evaporator bypass, high temperature fluid, an innovative control system and real-time visualisation on the respective diagrams (graphs not shown in this article). Furthermore, other advanced heat pump types could be integrated and tested. The focus is on convective dryers because they are most frequently used and can be conveniently modified to integrate HPs. By adapting HPs to operate under varying temperatures (by using fluid mixtures, hybrid and multi-stage HPs, HPs with variable pressure ratio etc) and developing innovative control systems, their efficiency can be increased substantially.
The test rig is designed to be transportable, with the heat pump including the condenser and evaporator mounted on a single frame. This allows demonstrations or research activities on other locations. Additional equipment such as heat exchangers and an evaporator bypass fan are installed as needed and placed in a rack.
To get the cycle architecture as flexible as possible, all heat exchangers can be (internally or externally) bypassed and all flows can be controlled. The heat pump features an open compressor, which enables precise mechanical power measurement and factors out compressor motor efficiency. In turn, this allows practical simulation of a gas engine powered heat pump dryer. Heat that would be recovered from the gas engine and exhaust at a given mechanical output, can be added as electrical heat injected after the condenser, by using data and extensive experience from CHP units. To aid repeatability and optimisation of a climate-dependent control algorithm, the dryer is fed with conditioned air capable of following a preset summer/winter day/night cycle.
Changes in humidity and enthalpy between ingoing and outgoing air of the drying chamber show that drying has taken place. The evaporation of moisture from the product is responsible for a rise in humidity, whereas an enthalpy (total heat content of a system) drop is caused by thermal losses from the drying chamber walls and heating up of the product. To reproduce conditions of ingoing and outgoing air in a typical drying process, evaporation followed by an air-to-air heat exchanger to create heat losses are included (see Fig. 4). This way, the heat pump dryer can be tested for a wide range of different drying schemes and products with corresponding temperature ranges, allowing simulation of batch and continuous processes from ongoing case studies. To our knowledge, a similar prototype has not yet been developed.
System Configurations and Process Simulations
To aid dimensioning of the test bench, a basic process simulation has been made with Matlab, making use of the Coolprop thermodynamic and psychrometric property libraries for fluids and humid air. With this model, state points and performance parameters of the drying and heat pump cycle are calculated using the following inputs (among others):
- System configuration and ambient conditions;
- Moisture extraction rate, fresh air ratio and evaporator bypass air ratio;
- Air temperature after the evaporator or after the condenser;
- Desired dryer inlet temperature;
- Relative humidity at dryer outlet;
- Primary energy ratio.
(1) Open-cycle dryer
So far, pressure losses, fan power input and drying mechanics have not taken into account. As a benchmark to compare the energy efficiency of more advanced dryers with, the basic open-cycle conventional dryer system configuration was simulated (shown in Fig 5). Additions are made in subsequent configurations to show possible optimisations. The efficiency of the heater is fixed at 90% and the drying process is considered to be isenthalpic, with a dryer outlet relative humidity of 55%. With ambient conditions of 18°C and 50% relative humidity, this results in a specific energy consumption (ε) of 4198 kJ/kg.
(2) Partially closed-cycle dryer
One of the most widely used ways to save energy in a drying process is to partially recirculate outgoing air. The addition of a bypass fan (shown in Fig. 7) cools the air that has passed through the humidifier. Having a higher temperature than ambient air, less energy is needed to heat it up to the desired temperature, but a higher moisture content means the driving potential of moisture transfer from product to air (fed into the humidifier) diminishes. If the other parameters remain constant, recirculating air leads to a higher moisture content in the dryer and thus increases efficiency at the cost of dryer capacity.
When calculating the energetic benefits by recirculation, the absolute humidity difference of the drying chamber was kept constant. By close approximation at these temperatures, this results in a rise in the dryer outlet relative humidity from 55% to 75%. With recirculation is set at 50%, energy consumption (ε) is decreased from 4198 kJ/kg to 3702 kJ/kg. Correspondingly, the two cycles require respectively 11.7kW and 10.4kW for a moisture extraction rate of 10kg/h. The test rig allows dynamic experimental optimisation of the recirculation ratio for different drying conditions and conditions of ingoing and outgoing air.
(3) Partially closed-cycle dryer with heat exchanger
A heat exchanger between ingoing and outgoing air offers clear advantages to the energy efficiency of a dryer, especially for high temperature drying applications. The ingoing air is preheated by cooling down outgoing air, possibly condensing water vapour. To keep the model simple, the heat exchanger effectiveness, defined as the ratio of the temperature rise of the ingoing air compared to the potential temperature rise, is set to 70%. This results in a ε of 3113 kJ/kg, down from 3702 kJ/kg in the previous example.
(4) Heat pump dryer with evaporator bypass and external heat exchanger
A heat pump can be implemented in the previously shown system as shown in Fig. 11. Whereas an air-to-air heat exchanger cannot transfer heat across the pinch point, a heat pump enables heat transfer from lower to higher temperatures. The selected heat pump fluid is R245fa, a HFC fluid with very similar characteristics as R134a but with higher critical temperature and saturation temperatures related to fluid pressure. Moreover, compared to R134a, the theoretical COP* is consistently slightly higher, with lower compressor discharge temperatures as an additional advantage. The fluid is fully compatible with standard commercially available ^HFC lubricants and components, respecting the normal limits for temperatures and pressures.
*COP (coefficient of performance) for a heat pump is the ratio of the energy transferred for heating to the input electric energy used in the process
^ According to Greenpeace, HFCs (hydrofluorocarbons)are the man-made greenhouse gases developed by the chemical industry to replace CFCs, HFCs’ ozone-killing cousins, which were banned in 1992 by the Montreal Protocol. HFCs are widely used in refrigeration, air conditioning, foam blowing agents, aerosols, fire protection and solvents, with refrigeration and air conditioning being major sectors of application. 134a is the most commonly used HFC.
In the simulation, compressor isentropic efficiency is set to 70% and 8K is chosen as pinch points of condenser and evaporator. Superheat is set to 10K and subcooling is negligible, because in steady state conditions with a vessel that is not entirely full, liquid enters and leaves the vessel at saturation conditions. In this example, the energetic optimum is reached when the saturation temperatures of the heat pump are 20 and 69.5°C. This results in a specific energy consumption ε of 1656 kJ/kg and 2380 kJ/kg if a primary energy ratio of 2.5 is taken into account (referring to a commonly used 40% average conversion efficiency of electricity generation). The external heat exchanger has a much smaller impact on the system as the evaporator cools down the outgoing air close to ambient temperature.
(5) Heat pump dryer with evaporator bypass, external heat exchanger and recirculation
As with conventional drying, the air can be (partially) recirculated leading to a (partially) closed cycle (see Fig. 14). In most cases, the energetic optimum for medium/high temperature drying is a more closed cycle, whereas low/medium temperature drying favours a cycle with more fresh air. Note that a fully closed cycle with a near-isentropic drying process requires an external condenser to dissipate excess heat as the condenser has a higher thermal capacity than the evaporator because of compressor heat. While the dryer inlet temperature is fixed, different cycles may yield different dryer inlet humidities. To retain dryer capacity, this would require additional fans in the drying chamber for increased air speed and/or possibly additional configuration changes depending on the product. As there is no submodel for drying mechanics at this point, outlet humidity will remain constant regardless of air flow.
The ideal evaporator pressure is increased when only 25% fresh air is being introduced in the cycle. As the air cycle shifts to a higher absolute humidity, the air going to the evaporator becomes a better heat source. This results in a ε of 1135 kJ/kg and 2053 kJ/kg if the primary energy ratio is taken into account. Maintaining the higher pressure ratio from Fig. 13 would result in too much heat extraction and could result in a too high temperature after the condenser. With this lower pressure ratio, a higher COP is obtained.
(6) Heat pump dryer with evaporator bypass, internal/external heat exchangers and recirculation
The principles of pinch technology state that as much as possible heat should be recovered by heat exchangers before using a heat pump. By employing an internal heat exchanger, as shown in Fig. 17, the air is pre-cooled before it goes to the evaporator and air is preheated before it goes to the condenser, decreasing the heat load and therefore the size of the heat pump. However, the resulting temperature difference between heat sink and source is increased and also leads to a decrease in COP. By experiments, it should be investigated if the pinch rules are still to be followed or a deviating energetic optimum can be found. A possible reason for this is a decrease in isentropic efficiency as pressure ratio rises and suction gas density decreases.
The internal HX also has its effectiveness set to 70%. This results in a ε of 1046 kJ/kg and 1914 kJ/kg if the primary energy ratio is taken into account.
In this case, the internal HX (heat exchanger) takes away all the sensible heat from the outgoing air of the dryer. This leads to a much smaller air temperature gradient over the evaporator and limits the ability to obtain superheated fluid at the end of the evaporator. To avoid shifting of the fluid-side pinch point from the beginning to the end of the evaporator, thus decreasing the evaporator pressure, superheat is limited to 5K instead of 10K. When applying this technique to a practical set-up, precaution is advised in the form of a liquid separator before the compressor. The addition of a suction line heat exchanger (SLHX) is presented in Fig. 20. In this configuration, heat is transferred from the high temperature liquid leaving the condenser to the low temperature refrigerant vapour leaving the evaporator. Thus, liquid refrigerant is subcooled before entering the expansion valve while vapour refrigerant is superheated before entering the compressor. Besides energetic benefits, a higher degree of subcooling leads to a smaller amount of flash gas after expansion and improves the evaporator filling grade and heat transfer. Additionally, using a SLHX helps to prevent flash gas formation at the expansion valve inlet and the risk of liquid refrigerant at the compressor inlet.
SLHX effectiveness is set to 70%. Optimum pressure ratio is slightly increased as heat pump performance improves. This results in a specific energy consumption ε of 728 kJ/kg and 1688 kJ/kg if the primary energy ratio is taken into account.
Depending on the superheat at the outlet of the evaporator, effectiveness of the SLHX and temperature difference between condensing and evaporating sides, suction gas superheat might become too elevated. High superheat causes high compressor discharge temperatures and lower isentropic efficiency. This lower isentropic efficiency is due to a decrease of suction gas density, requiring a higher compressor speed despite a lower mass flow. For this reason the SLHX of the test bench is equipped with a 3-way proportional bypass valve to decrease the effectiveness of the SLHX if needed. Desuperheating with fluid injection can be implemented as well. Fig. 23 presents the use of a fluid-to-air subcooling heat exchanger. A subcooling heat exchanger offers a substantial improvement to the COP of a single-stage heat pump when there is a large heat sink temperature gradient. Having a similar application range as CO₂ heat pumps, simulations suggest subcooling R245fa systems may prove to be an alternative that allows low heat sink temperature gradients without other usual limitations of a CO₂ system. Notably, the heat source temperature of a CO₂ system shouldn’t be too high to prevent supercritical conditions in the evaporator.
The simulated subcooler has a surface area of 35% compared to the condenser, where the assumed heat transfer potential is based solely on the LMTD (log mean temperature difference). Performance with a crossflow, non-ideal heat exchanger is expected to be worse than the simulation suggests, so a heat exchanger with a surface ratio of 50% was selected for the practical setup. Optimum pressure ratio is slightly increased as heat pump performance improves. This results in a ε of 637 kJ/kg and 1528 kJ/kg if the primary energy ratio is taken into account.
Putting Possibilities Into Practice
This paper presents the possibilities of a heat pump dryer test bench under construction with the help of a simulation model. Once the model is optimised and validated by tests, it will enable the design of energy-efficient drying systems by either using an existing installation or starting from scratch. These possible concepts also include the use of a gas engine driven heat pump (favoured in regions where electricity is pricier), which has yet to be included in the model. Simulations reveal R245fa to be an all-round well performing fluid for heat pump dryers, especially for the higher temperature range. Depending on heat sink and source conditions, significant energetic benefits can be realised with a SLHX and subcooler. By utilising an internal heat exchanger in the air cycle, the required heat pump capacity decreases. Furthermore, the amount of heat extracted can be controlled not only by the heat pump pressure ratio, but can be complemented by an evaporator bypass. Besides demonstrating optimisation possibilities on the drying cycle, the test rig will offer an excellent environment for research on advanced heat pump systems including novel control systems.