System cost of energy generation scenarios for South Africa: understanding the real cost of integrating energy generation technologies

Sklar-Chik, Mark David (2017-03)

Thesis (MEng)--Stellenbosch University, 2017.

Thesis

ENGLISH ABSTRACT: South Africa’s electricity market has realized much growth with the introduction of the Renewable Energy Independent Power Producers Program (REIPPPP). This program enables the increase in energy generation from renewable technologies. Moving forward, the future targets for renewable generation increase to levels where they contribute significantly to system supply. This research sought to understand the system costs of integrating utility scale renewable energy generation technologies into the South African electricity system and thus considered future energy scenarios with higher renewable penetration levels. In addition, the research investigated Levelized Cost of Energy for the generating technologies encompassing renewables (solar PV, CSP and wind) and non-renewables (coal, gas nuclear). A search through literature exhibited many electricity-modelling tools. A bottom-up approach was chosen, which captured the technical details of the generators and electricity network. Further, the selected electricity modelling software was PLEXOS. This tool enabled the capture of the South African electricity network, including all the generators. The model was a single node model, where the system demand aggregates at the node. Using the targets of a WWF high scenario, where renewables’ penetration was 25% by energy and 41% by installed power, an hourly simulation was run for the year 2030, while 2010 actual system demand was used and forecasted to 2030. The attempt was to understand the real system cost. Hence, a base scenario with unconstrained generators and then a constraints scenario containing all generator parameters such as Minimum Stable Level, availabilities and so forth, was run. The results showed the base case had a system cost of R0.39/kWh, while the constraints scenario R0.48/kWh. In both scenarios, the unserved energy was negligible compared to the total generation costs. The renewable energy total capacity factor was 29% for the simulation. Total generation for the year was 409819.07 GWh and the corresponding total cost was 10 trillion Rand. From the constraint model, the LCOE for CSP was R1.44/kWh, second was solar PV at 1.25/kWh, and wind was R1.02/kWh, while the lowest were the existing plants (OCGT, hydro, nuclear, pumped storage) well below R0.65/kWh, as their capital and interest were assumed to have been settled by the start of the simulation. Integration elements comprising the number of generators’ start-up and shutdown, water consumption and emissions were quantified. The emissions were significant cost contributors, when using the price of R48/tonne. In the sensitivity analysis, the following input parameters were tested: fuel price, generator availabilities, system demand, and increase in renewable energy production. Dropping the system demand to WWF low levels affected the system cost the most, increasing the value to R0.59/kWh. Whereas increase of the renewables production profile of 10% caused the system cost to drop to R0.52. This showed that the demand forecast is crucial for modelling system behaviour. The research fulfilled the objective and demonstrated the system costs of integrating renewables into future energy scenarios. Future models should include transmission and distribution infrastructure, more detailed generator performance criteria. The conversion from solar or wind resource to renewables output plants must be further investigated. Additional recent costing data, updated demand forecasts and smaller non-utility scale projects should be incorporated in future models.

AFRIKAANSE OPSOMMING: Suid Afrika se elektrisiteit mark het baie groei ervaar met die bekendstelling van die Renewable Energy Independent Power Producers Program (REIPPPP). Hierdie program maak dit moontlik vir 'n toename in krag-opwekking deur hernubare tegnologieë om plaas te vind. Die toekomstige teikens vir hernubare krag-opwekking sal toeneem tot vlakke waar hulle 'n merkwaardige bydrae sal lewer tot die krag-toevoer stelsel. Hierdie navorsingstuk beoog om die stelsel-kostes te verstaan wat geassosieer word met die integrasie van nut-skaal hernubare energie tegnologieë in die bestaande Suid Afrikaanse elektrisiteit toevoer stelsel deur die ondersoek van moontlike toekomstige energie scenario's met groter bydraes deur hernubare energie. Boonop, het hierdie studie ten doel om die Levelized Cost of Energy (Verdeelde Lewens Koste van Energie) van die voorgestelde hernubare energie tegnologieë, as deel van REIPPPP, en die van nie-hernubare tegnologieë, in die verskillende scenario's, te ondersoek. 'n Inspeksie van die literatuur het vele elektrisiteit modellerings pakkette uitgelig. 'n Sogenaamde "bottom-up: benadering was gevolg wat die tegniese besonderhede van die opwekkers en netwerk vasgelê het. Die gekose modellerings pakket was PLEXOS, wat die uiteensetting van die Suid Afrikaanse elektrisiteits netwerk moontlik gemaak het. Hierdie model is 'n enkel-node model, waar die stelsel-aanvraag by 'n node saamgevoeg word. Die teikens uiteengesit in die WWF hoë-geval was gebruik, met hernubare energie bydraes van 25% vir jaarlikse energie en 41% geïnstalleerde kapasiteit. 'n Uurlikse simulasie is toe opgestel vir die jaar van 2030, met 2010 se jaarlikse aanvraag as die verwysingspunt en vooruit geskat tot 2030. Die uitgangspunt was om die werklike stelsel kostes van so moontlike toekomstige geval te verstaan. In lig hiervan was 'n basis geval ook gesimuleer, met onbeperkte opwekkers-parameters, sowel as n geval met beperkte inset-parameters soos minimum stabiele vlak en beskikbaarheid, onder andere. Die resultate het getoon dat die basis geval 'n stelsel koste van 0.39 R/kWh bereik, terwyl die beperkte geval 'n koste van 0.48 R/kWh het. In beide gevalle was die ongedienste energie weglaatbaar klein in vergelyking met die opwekkings kostes. Die totale kapasiteits faktor van die hernubare energie bydraes was 29% vir die simulasie. Die totale energie opgewek vir die jaar was 409819.07 GWh met 'n ooreenstemmende totale koste van 10 triljoen Rand. Die onbeperkte model het 'n LCOE waarde van 1.44 R/kWh vir CSP gelewer, terwyl wind 1.02 R/kWh, en PV 1.44 R/kWh was. Die laagste LCOE was bereik deur die bestaande vloot opwekkings eenhede (OCGT, hidro-, kern- en gepomp-stoor hidrokrag) teen 0.65 R/kWh, aangesien dit aangeneem is dat hul kapitaal en rente reeds afbetaal is teen die begin van die simulasie. Integrasie elemente was gekwantifiseer, wat bestaan uit die aantal opwekker-eenheid aan-skakelings en af-skakelings, water verbruik en afval-gas vrystellings. Die afval-gas vrystellings het merkwaardig bygedra tot die kostes teen 'n prys van R48 per ton. In die sensitiwiteits analise was die volgende inset-parameters getoets: brandstof prys, opwekker beskikbaarheid, stelsel-aanvraag en toename in hernubare energie produksie hoeveelhede. Dit was gevind dat indien die stelsel aanvraag verlaag was tot die WWF lae-geval vlakke, dit die stelsel koste die mees beïnvloed het, met 'n verhoging tot 0.59 R/kWh. Aangesien toename van die volhoubare produksie profiel van 10% het veroorsaak dat die stelsel koste te daal tot R0.52. Dit wys duidelik dat die aanvraag vooruitskatting van kardinale belang is vir deeglike stelsel modellering. Dit was dus gevind dat die navorsingstuk die uiteengesette doel vervul het die stelsel kostes van nut-skaal hernubare energie integrasie uiteengesit. Soortgelyke toekomstige modelle sal klem moet lê op transmissie en verspreidings infrastruktuur, en in meer detail kyk na opwekker werkverrigting. Die oorgang van sonk- of wind hulpbronne tot hernubare energie opwekkers moet vêrder nagevors word. Addisionele, meer onlangse koste data, vanaf opgedateerde aanvraag vooruitskattings en kleiner nie-nut skaal projekte moet ook geïnkorporeer word in toekomstige modelle.

Please refer to this item in SUNScholar by using the following persistent URL: http://hdl.handle.net/10019.1/100811
This item appears in the following collections: