Absorption based seasonal thermal storage with sodium hydroxide, Progress and Outlook

Transcription

1 19. Status-Seminar «Forschen für den Bau im Kontext von Energie und Umwelt» Absorption based seasonal thermal storage with sodium hydroxide, Progress and Outlook B. Fumey a (benjamin.fumey@empa.ch), R. Weber a (robert.weber@empa.ch), P. Gantenbein b (paul.gantenbein@spf.ch), X. Daguenet-Frick b (xavier.daguenet@spf.ch), L. Baldini a (luca.baldini@empa.ch) a: Empa, Urban Energy Systems Laboratory, Überlandstasse 129, CH-8600 Dübendorf b: Institute for Solar Technology SPF University for Applied Sciences-HSR, Oberseestr. 10, CH-8640 Rapperswil, Switzerland Zusammenfassung Résumé Abstract An essential part in the utilization of solar thermal power in order to enable renewable heating and seasonal load shifting is seasonal storage. In the past 40 years plus, numerous developments have been made to achieve this. Initially most activities focused on storing sensible heat in a diversity of materials and shapes, taking into account the high heat loss and large volume requirements associated. In the scope of the International Energy Agency (IEA) Solar Heating and Cooling Program (SHC) Tasks 32 [1] and 42 [2] new approaches have been studied. Sorption and chemical type heat storage technologies are proposed to reach both reduction in heat loss during storage time and increased energy density. This work remains an on going challenge today. Advanced heat storage technologies such as absorption heat storage enable potentially lossless storage, when not including charging and discharging processes, and show prospective for increased volumetric energy density in comparison to sensible hot water storage. In absorption heat storage not sensible heat is stored, but the potential to regain heat. In the framework of the Swiss Federal Office of Energy (SFOE) funded project NaOH-Speicher für saisonale Wärmespeicherung, and the EU funded project COMTES research and development has been done at Empa in collaboration with HSR-SPF in the field of heat storage based on the absorption of water vapour on sodium hydroxide. In both projects, demonstrator type setups were built in order to gain experience on system and component performance, detect challenges and identify performance gaps. The key component of said technology is the heat and mass exchanger, interconnected to respective working fluid storage tanks. In the two projects varying approaches were followed. In the SFOE project a plate type heat and mass exchanger version with a double floor and in the EU project a tube bundle falling film version was built. Operation of both units was challenging and large performance gaps with reference to expected performance were identified. From this work, it is concluded, that novel heat and mass exchanger types are required in order to enable continuous stable high efficiency performance. Continuation of this work at Empa and HSR-SPF, is planned in conjunction with the SCCER Heat and Electricity Storage program as well as imbedded in a new IEA SHC Task. Further work will focus on novel and improved heat and mass exchangers along with in depth analysis of performance gaps leading to improved system performance and increased technology readiness level for potential market entry. 8. / 9. September 2016 ETH-Zürich 1

2 1. Scope Switzerland s aim for a 2000 W per capita society by 2050 [3] represents a drastic 63 % reduction in the current level of energy consumption. Up to date 60 % (equivalent to 3000 W per capita) of Switzerland s energy demand is met by fossil fuels, principally oil and natural gas. The building sector covers approximately 50 % of the total consumed final energy, whereby a third of this is used for space heating. There remains an urgent need to rapidly move towards the harvest and utilisation of renewable energy. To this objective, storage plays an important role in synchronising energy demand to fluctuating supply, both on a short and long term or seasonal basis. In respect to building energy consumption, the Swiss Application Purpose Report 2014 [4] reveals that 71 % (27.1 TWh) of all private home areas and 59 % (5.2 TWh) of all domestic hot water in Switzerland is still heated from fossil fuel. Due to a lack of compact, long term thermal storage, the use of solar thermal energy to this purpose remains marginal. The last 40 years have seen many variations of sensible heat based seasonal thermal energy storage systems developed. These have been based on rock beds, caverns, tanks, etc. under ground and above ground. The use of hot water tanks for seasonal storage is one of the best known thermal energy storage technologies and has been established on the market as an affordable solution for new low energy buildings. Using this technology, solar thermal energy is stored in large well insulated water vessels. Nevertheless, due to the large volume required and the constant thermal loss annual coverage is often limited. When bridging longer supply gaps storage temperatures may not be sufficient for DHW. Even so, this technology shows the high potential for solar heating, whereby shortcomings highlight the need for more research towards compact and lossless thermal storage. Work in this field has been on going at Empa and in part in collaboration with HSR-SPF. Absorption based heat storage shows potential for lossless storage and volumetric energy capacity increase of up to factor 5 compared to the sensible thermal water storage system. The Empa developed absorption heat storage system is based on the absorption of water, referred to as absorbate, on aqueous NaOH, referred to as absorbent. The concept likens a thermally driven absorption heat pump [5] extended with absorbent and absorbate storage. The main system components are, absorbent and adsorbate storage tanks and a heat and mass exchanger (HMX). Both water and aqueous NaOH are stored in their liquid state and can thus be pumped to and from the HMX and tanks. The HMX consists of a unit operating as absorber and desorber interconnected to a unit serving as evaporator and condenser respectively. The process functions under the exclusion of all gasses apart from water vapour, at the respective temperature dependent water vapour pressure [6]. In practice this requires vacuum compliant components [7]. The working principle is as follows: In the charging process, sodium hydroxide (NaOH) solution with high water content is heated from a solar thermal source and water is partially evaporated. The water vapour is condensed on the condenser, whereby the heat of condensation is released to a respective heat sink. In this way, heat is not stored in its sensible form, but the potential to regain heat at elevated temperatures from a low temperature heat source is stored. No loss in this stored potential is encountered, in as far as mixing of the aqueous NaOH solution and the separated water is prevented. In order to regain heat at elevated temperatures, water is evaporated on the evaporator by sourcing a low temperature heat supply such as a vertical bore hole heat exchanger. The resulting water vapour is absorbed on the aqueous NaOH solution with low water content, whereby it is liquefied, and the heat of condensation is released to the NaOH [5]. Due to the high affinity of aqueous NaOH to water, depending on the NaOH content, a respective temperature increase is reached. In the approach, concentration of 50 wt% NaOH in water in its charged state is reached [6]. This upper limitation was set due to the solidification of aqueous NaOH at higher concentrations under room temperature, whereby the solution transport by pumping is prevented [5]. The resulting theoretical maximum temperature gain is 38 K [8] whereby temperature drops over the HMXs are expected to lead to a lower actual gain of approximately 32 K. The absorption heat storage output depends on the temperature of the evaporator and the sorbent solution concentration. The storage energy density further depends on the minimum temperature difference between evaporator and absorber. Reduced temperature difference permits increased water vapour uptake and thus improves system energy density. It can be noted that the system output temperature and its energy density are highly dependent on operating parameters [8]. In the following text the developed HMXs for the two different projects are described, results discussed and an outlook towards further development is given. Benjamin Fumey 19. Status-Seminar 8. / 9. September 2016 ETH-Zürich 2

3 2. SFOE project: NaOH-Speicher für saisonale Wärmespeicherung The objective of this project was to develop a lab scale demonstrator of a seasonal thermal storage based on the absorption of water vapour on NaOH as described in the previous section. This demonstrator system should allow heat supply for space heating (SH) and domestic hot water (DHW) in a low energy house. To this purpose, a plate type HMX was developed as illustrated in figure 1. Due to the non full cycle operation of the system, charging in summer and discharging in winter, rather than constructing four HMXs operating as absorber, desorber, evaporator and condenser respectively, only two were built. These serve as absorber and desorber (AD) as well as evaporator and condenser (EC) respectively. In operation, the plates were filled with NaOH, whereby this slowly flowed from one plate to the consecutive plate below. The heat transport fluid flowed in counter flow between the double walled plate bottom as visualised in figure 2. In this setup a batch type continuous process was possible. Batch type, due to the collection of aqueous NaOH on the plates, and continuous due to the flow of NaOH from one plate to the consecutive. Both AD as well as EC unit were built identically, whereby the EC unit consists of 4 plates and the AD unit holds 3 plates. Heat transport fluid OUT Aqueous NaOH IN Aqueous NaOH OUT Heat transport fluid IN Figure 1 Illustration of the aqueous NaOH and heat transport fluid in the plate type double floor HMX. In desorption testing, the AD unit was preheated to approximately 78 C. In the test sequence shown in figure 2, the EC unit is not cooled during preheating. For this reason, up to time 108 min, the temperature on C4 is higher than on the other EC plates, due to radiation from the AD unit plate D1. At time 108 min., noted as Start in the diagram, diluted NaOH was introduced to the desorber, and the condenser was cooled. The steep temperature drop points to the evaporation of water from the aqueous NaOH. Vapour is in turn condensed on the condenser, for this reason the condenser surface temperature is higher than the cooling water, which was set to 12 C. Benjamin Fumey 19. Status-Seminar 8. / 9. September 2016 ETH-Zürich 3

4 Temperature [ C] 80 D C Start Stop TC1 TC2 TC3 TC4 TD1 TD2 TD3 Time [min] TC1 Figure 2 TC2 TC3 TC4 TD3 TD2 TD1 The diagram shows the measurement sequence from a desorption test on the left, and the plate type HMX on the right. Through the plate type build, homogeneous distribution of the NaOH on the plate was difficult. This especially affected the absorption process. Fresh, charged NaOH was immediately mixed with partly discharged NaOH on the plate, thus temperature gain was strongly decreased. Continuous flow of the aqueous NaOH could not be guarantied. In order to cover the plates with NaOH and permit NaOH flow, extra NaOH was required on the plates, leading to increased distance from the absorption surface to the heat transport fluid, resulting in increased heat transfer resistance. The very bulky build of both the plates and the vacuum containment chamber, further lead to high heat losses. In order to prevent pressure drop in the absorbate vapour transport, both the AD and EC HMXs were contained in one vacuum vessel. The EC HMX was placed above the AD HMX in order to prevent aqueous NaOH from contaminating the condensed water if spilled. A uniform vapour flow from AD to EC or vice versa dependent on operation mode was assumed. Thus, no thermal loss from circulating vapour was expected. Nevertheless, in operation it was found that placing both AD and EC HMXs in the same vessel lead to high direct thermal transport between the two HMXs via radiation and water vapour circulation. EU 7th framework program project COMTES In the COMTES project a full scale heat storage demonstrator for a single family home was built. To this purpose the falling film tube bundle HMX design was chosen [9]. This permits compact geometric design, high volumetric power density [10], smooth continuous operating process and large contact area of absorbent to absorbate and heat exchanger as well as absorbent turbulence for higher heat transfer rates. Figure 3 shows the EC unit of the HMX. Benjamin Fumey 19. Status-Seminar 8. / 9. September 2016 ETH-Zürich 4

5 Figure 3 CAD drawing of the reaction zone with both A/D HMX unit on the left and E/C HMX unit on the right. As in the initial plate type HMX, absorber and desorber as well as evaporator and condenser are combined respectively to an AD and an EC unit. This is done in order to increase power density and storage energy density. However, the challenge in the design of the components is to master the different heat and mass transfer rates in the process steps in one component. Both tube bundles are installed in separate chambers, interconnected for vapour exchange. A critical point in this design is the homogeneous distribution of the fluids on the outer tube surface in order to achieve high surface contact to the tubes as well as to the absorbate vapour, leading to an efficient heat and mass transfer. Special attention was given to the design of the manifold suspended over the tube bundle, performing the even distribution of absorbent and absorbate fluid over the respective tubes. As previously mentioned, mixing of fresh concentrated absorbent and diluted absorbent is not ideal due to the dependency of the temperature lift to the absorbent concentration. For this reason recirculation of the absorbent on the absorber as is common on tube bundle HMXs was not possible, making good wetting a challenge at high concentration and low flow rates. In order to dimension the AD and EC tube bundle HMXs, a mathematical model was established [11]. From this an AD unit with 4 x 18 tubes of 300 mm length and 10 mm outer diameter, and an EC unit with 16 x 12 tubes of 700 mm length and 10 mm outer diameter was built. The calculated power was 12 kw in desorption mode and 8 kw in the absorption process. Several absorption and desorption tests have been done with the HMX illustrated in figure 3. Results of two absorption tests are visualised in figure 4. These illustrations show steady state average values. The blue circles indicates the tube bundle arrays for both the AD and EC units. The heat transport medium flows upward from one tube to the other, as shown by the blue arrows through the circles. The orange arrow shows the flow of the absorbent in the AD unit, this is in counter flow to the heat transport fluid. In the EC unit the sorbate flow is indicated in dark blue. Temperatures are measured inside as well as outside of the tubes at varying levels, as indicated. Volume flows of the sorbent and the heat transport fluids in both AD and EC units are specified as well as the power input and output. In operation it was discovered that non condensing gases have a substantial effect on the absorption process. In the figure 4 on the left, operating at a pressure of 1.9 mbar above the respective water vapour pressure (18.7 mbar abs.), stable vapour transport equivalent to 1 kw of thermal energy could be reached. On the other hand on the right, at similar settings, the measured pressure was 10.3 mbar above the water vapour pressure (17.4 mbar abs.). In this test no vapour transport was possible. Benjamin Fumey 19. Status-Seminar 8. / 9. September 2016 ETH-Zürich 5

6 Figure 4 Illustration of the absorber / desorber (AD) and evaporator / condenser (EC) HMX test results. On the left is the result with nearly all non condensing gases removed, on the right a result with approximately 10 mbar pressure resulting from non condensing gases. Even though a continuous power output of approximately 1 kw was achieved, the actual temperature gain was only approximately 2 C, from C to C. Theoretically an output temperature of approximately 56 C should be possible [5], when neglecting temperature losses. The sorbent concentration reaches only a slight reduction, which explains the low temperature gain and power output. It appears that the conventional HMX as designed in the COMTES project and often used in absorption heat pumps [9] is not fitting to the heat storage application using aqueous NaOH. This can be explained as follows: Even though the process of absorption and desorption in the heat storage application is comparable to the solar thermal driven heat pump, the actual operation differs strongly. Solar chilling for example follows a continuous full cycle process. Charging and discharging are continuous processes, whereby the quantity of absorbate transported between desorption and absorption is not of primary importance. In the heat storage approach on the other hand a continuous but not full cycle process is at work. This is due to the nature of its role as storage. In the heat storage demonstrator, especially in the absorption process, it is of significant importance that a low concentration of absorbent in the solution, in other words high water vapour absorption is reached. This has direct impact on the storage energy density. Due to the dependence of the output temperature on the aqueous NaOH solution concentration, absorption must occur in one step and recirculation is not possible. Consequently the exposure time of the aqueous NaOH solution to the water vapour must be sufficient to reach high water uptake. In the COMTES setup an average residence time of about 2 seconds was recorded. A further issue is found in the formation of aqueous NaOH droplets. Figure 5 shows the AD tube bundle with the manifold evenly supplying small droplets on the first tube. Nevertheless they quickly group to form larger droplets, thus reducing the wetted area of the tube bundle and decreasing the fall through time, whereby reducing both contact area and contact time of aqueous NaOH to water vapour. Benjamin Fumey 19. Status-Seminar 8. / 9. September 2016 ETH-Zürich 6

7 Figure 5 Picture of the tube bundle in the AD unit, showing the flow of aqueous NaOH over the tubes. Fine droplets group, reducing the wetted surface on the tubes. 3. Conclusion and Outlook It is found that a novel HMX design for the absorption process in absorption heat storage is required. In the design it is critical that the aqueous NaOH is allowed adequate time for water vapour absorption. This could require a suspension in water vapor time of more than one hour. The HMX needs to be designed in a way that prevents regrouping of aqueous NaOH, in order to reach high surface wetting area for water vapour as well as heat exchange. New and alternative approaches to HMXs are now being developed and tested at Empa on a lab scale level. An up scaling of promising designs towards a replacement in the COMTES built hybrid thermal storage system will follow this work, after which the complete system will undergo further testing for concept and system validation. The continuation of this work at Empa and HSR-SPF will be carried out in the frame of the SCCER Heat and Electricity Storage program as well as within a forthcoming IEA Solar Heating and Cooling technology collaboration program Task. Focus is placed on novel and improved HMXs as well as supportive components, leading to improved system performance and increased technology readiness level for potential market entry. 4. Acknowledgements Financial support by the SFOE under the project No and the EU 7 th framework program under the grant agreement No is gratefully acknowledged. Furthermore, the authors gratefully acknowledge financial support of the research institutions Empa Swiss Federal Laboratories for Materials Science and Technology and HSR University of Applied Sciences Rapperswil as well as the industrial partner Kingspan environmental. References [1] IEA SHC Task 32 - Advanced Storage Concepts for Solar and Low Energy Buildings, July 1, 2003 to December 31, 2007 [2] W. van Helden, et al, IEA SHC Task 42 / ECES Annex 29 Working Group B: Applications of Compact Thermal Energy Storage SHC 2015, [3] Koschenz M, Pfeiffer A, Wokaun A. Energy and building technology for the 2000 W society - Potential of residential buildings in Switzerland. Energy and Buildings 2005; 37: [4] Analyse des schweizerischen Energieverbrauchs nach Verwendungszwecken, Oktober 2015 (German) Benjamin Fumey 19. Status-Seminar 8. / 9. September 2016 ETH-Zürich 7

8 [5] Benjamin Fumey, Robert Weber, Paul Gantenbein, Xavier Daguenet-Frick, Tommy Williamson, Viktor Dorer, Development of a Closed Sorption Heat Storage Prototype, Energy Procedia, Volume 46, 2014, Pages [6] B. Fumey, R. Weber, P. Gantenbein, X. Daguenet-Frick, T. Williamson, V. Dorer, Closed sorption heat storage based on aqueous sodium hydroxide, Energy Procedia Volume 48, 2014, Pages [7] Weber R, Dorer V. Long-term heat storage with NaOH. Vacuum 2008;82: doi: /j.vacuum [8] B. Fumey, R. Weber, P. Gantenbein, X. Daguenet-Frick, I. Hughes, V. Dorer, Limitations imposed on energy density of sorption materials in seasonal thermal storage systems, Energy Procedia Volume 70, 2015, Pages [9] X. Daguenet-Frick, P. Gantenbein, E. Frank, B. Fumey, R. Weber, T. Williamson, Reaction Zone Development for an Aqueous Sodium Hydroxide Seasonal Thermal Energy Storage, Energy Procedia, Volume 57, 2014, Pages [10 ] Roques JF, Thome JR. Falling Film Transitions Between Droplet, Column, and Sheet Flow Modes on a Vertical Array of Horizontal 19 FPI and 40 FPI Low-Finned Tubes. Heat Transf Eng 2003;24:40 5. doi: / [11] X. Daguenet-Frick, P. Gantenbein, E. Frank, B. Fumey, R. Weber, Development of a numerical model for the reaction zone design of an aqueous sodium hydroxide seasonal thermal energy storage, Solar Energy, Volume 121, 2015, Pages Benjamin Fumey 19. Status-Seminar 8. / 9. September 2016 ETH-Zürich 8