Design and development of a thermal rock bed storage experimental facility

Erasmus, Jeannie (2019-04)

Thesis (MA)--Stellenbosch University, 2019.

Thesis

ENGLISH ABSTRACT: The value of concentrating solar power (CSP) plants lies in dispatchability, which is provided through an integrated cost-effective thermal energy storage system (TESS). Compared to current state of the art molten salt thermal energy storage systems, a rock bed thermal energy storage system has the potential to reduce both the capital costs and Levelized Cost of Electricity (LCOE) significantly. The Stellenbosch University (SU) first-generation rock bed thermal energy storage design served as a proof of concept while the second-generation rock bed design was designed for significant cost reduction. This work presents the third-generation rock bed TESS at SU, through partial re-design, predominantly aiming at maximizing the usable rock mass. The rock bed thermal energy storage system is charged by air at a temperature of 650 C. An existing experimental facility, based on the second-generation design, has recently been constructed. To modify the facility, a concept was developed with knowledge gathered from both the first and second-generation concepts. The new concept charges the rock bed from the top downwards, with a predicted near-linear thermocline progression, where the thermocline is defined as the transition layer from the high temperature to the low temperature within the rock bed. Although the concept has a higher capital cost, an improved performance is predicted for the entire system. After development, the concept was adapted to the existing facility. Three experimental test campaigns were conducted, concluding with a multiple cycle test. This test consisted of three charge-discharge cycles, where the rock bed was discharged to a minimum outlet temperature of 327 C. Determining an accurate discharge mass flow rate was a challenge throughout testing, with flow leakages detected within the system. A flow loss assumption of 40 % was made after several cold air flow rates were tested. The second cycle within the multiple cycle test yielded a heating capacity of 336.67 kWhth, a volumetric efficiency of 60.30 % and a thermal efficiency of 92.40 %. An overall efficiency of 94.24 % was achieved over the three cycles. An analytical model was developed to be validated by the experimental results. From the validation, a possible prediction can be made on the performance of such a rock bed thermal energy storage system on an industrial scale. The thermal efficiency comparison yielded a maximum difference between the experimental and analytical results of +8.00 % for the first two cycles and +19.36 % for the third cycle. It is clear from this comparison that the model over-predicts the performance of the facility. Considering that the model is one dimensional and that it disregards both radiation and convection as heat transfer elements, as well as thermal losses, the model appears to be acceptable. However, it is recommended that further improvements be made to the model for a more accurate comparison. The overall results show that there has been an improvement in performance of the rock bed after the design changes that were made. These design changes include the addition of insulation and introducing the high temperature air into the top of the rock bed, rather than at the bottom. Room for improvement on the design to achieve higher overall performance has been identified and possible solutions are presented within this project.

AFRIKAANSE OPSOMMING: Die waarde van gekonsentreerde sonkragaanlegte lê in versendbaarheid, wat deur n geïntegreerde koste-effektiewe termiese energie stoorstelsel voorsien word. n Termiese energie stoorstelsel wat bestaan uit n klipbed, het die potensiaal om die kapitaalkoste aansienlik te verminder in vergelyking met die huide gesofistikeerde sout termiese energie stoorstelsels. Die tweede-generasie klipbed was ontwerp vir beduidende kostevermindering. Hierdie projek lê die derde-generasie klipbed termiese energie stoorstelsel by Stellenbosch Universiteit voor, en fokus op die gedeeltelike her-ontwerp van die tweede-generasie stelsel, wat hoofsaaklik daarop gemik is om bruikbare klipmassa te maksimeer. Die klipbed termiese energie stoorstelsel maak gebruik van lug teen n temperatuur van 650 C as hitteruilings vloeistof. n Ontwikkelde eksperimentele fasiliteit, gebaseer op die tweede-generasie ontwerp, bestaan reeds. Elemente van beide die eerste en tweede-generasie ontwerpe was in ag geneem om ’n verbeterde konsep te ontwerp. Die nuwe konsep laai die klipbed van bo na onder, met ’n voorspelde naby-lineêre termofront progressie. Alhoewel die konsep n hoër kapitaalkoste het, word ’n verbeterde prestasie vir die algehele stelsel voorspel. Na die ontwerpfase is die konsep aangepas tot die bestaande eksperimentele fasiliteit. Na afloop van die konstruksie was drie experimentele toetse uitgevoer, met n meervoudige siklus toets die laaste en belangrikste. Hierdie toets het bestaan uit drie laai-ontlaai siklusse, waar die stelsel elke keer ontlaai was tot n minimum uitlaat temperatuur van 327 . Die bepaling van n akkurate massa vloeitempo was n uitdaging gedurende die experimentele toetse as gevolg van vloei lekkasies wat plaasgevind het in die sisteem. n Vloei verlies van 40 % was gebruik as aanname na verskeie koue lug vloeitempo toetse. Die tweede siklus van die toets het n verhittingskapasiteit van 336.67 kWhth gelewer, tesame met n volumetriese effektiwiteit van 60.30 % en ’n termiese effektiwiteit van 92.40 %. Die meervoudige siklus toets het n algehele doeltreffenheid van 94.24 % bereik. n Analitiese model was ontwikkel vir die moontlike voorspelling rakende die prestasie van die nuwe konsep op n industriële skaal. Die eksperimentele resultate was gebruik om die model te verifiër, deur die resultate te vergelyk. Die vergelyking het ’n maksimum verskil van +8.00 % gelewer na die analitiese model se rigting vir die eerste twee siklusse en +19.36 % vir die derde siklus. Hierdie verskil was tussen die experimentele resultate en analitiese model se termiese doeltreffenheid. Indien dit in ag geneem word dat die model slegs een-dimensioneel is en beide konveksie en radiasie ignoreer as hitteoordrag elemente, sowel as hitteverlies na buite elemente, is die model aanvaarbaar. Daar word wel aanbeveel dat verdere verbeteringe aangebring word tot die model om n meer akkurate vergelyking te tref. Die algehele resultate toon dat daar n definitiewe verbetering is in terme van die prestasie van die stoorstelsel na die veranderinge wat plaasgevind het. Die toevoeging van insulasie en die laai van die klipbed van bo word gesien as positiewe veranderinge. Daar is egter nog baie ruimte vir verbetering om selfs hoër doeltreffenhede te bereik. Aanbevelings om die stoorstelsel te verbeter word binne hierdie projek gemaak.

Please refer to this item in SUNScholar by using the following persistent URL: http://hdl.handle.net/10019.1/105862
This item appears in the following collections: