The current bid selection criteria of the Renewable Energy Independent Power Producer Procurement (REIPPP) programme’s narrow evaluation and spatial blindness favours projects that are not necessarily the least-cost option for renewable energy at the point of consumption. This not only impacts energy costs for Eskom, and therefore for consumers, but also creates transmission congestion and delayed connections to an already constrained grid, incurring further costs.
Renewable energy (RE) generation forms part of the national integrated resource plan for electricity, IRP 2010-2030, a plan adopted by the DoE in 2011 that establishes an energy generation mix and profile for South Africa up to 2030. The REIPPP programme is the implementation vehicle for RE in terms of this plan.
IRP 2010-2030 allocates 17 800MW for RE generation, which includes solar photo-voltaic, concentrating solar, wind, gas and hydro power generation. Although some generation sources are geographically bound (e.g. coal), others such as solar PV are not.
Solar PV best exemplifies the bid criteria shortfalls for two reasons: solar PV forms a large part of the REIPPP programme, and solar PV is more “flexible” in respect of geographic location in that solar irradiation is good in most parts of South Africa.
Fig. 1: Solar PV makes up a large proportion of planned RE generation capacity.
Most of the planned 8 400MW of IPP solar PV farms allocated under the IRP 2010-2030 have to date been placed in the “Solar Corridor” – an area in the Northern Cape where solar irradiation is optimal. Currently, some 6 400MW is allocated in the Solar Corridor, with another 2 000MW distributed elsewhere in the country.
Flawed bid selection criteria
For the procurement, the South African Department of Energy (DoE) favoured a “competitive tender” bidding approach over a “feed-in-tariff” (the latter being a widely used international policy instrument for procuring RE).
The current bid evaluation process has three parts, the third and most important part being the comparative evaluation.
In the comparative evaluation, the levelised cost of RE supplied at the point of connection to the grid (R/kWh) and the socio-economic development objectives are the determining factors in the comparative evaluation, and are weighted 70:30 respectively.
Fig. 2: Contractual arrangement for RE generation. [Credit: Eskom]
There are five bid phases (or windows) – the first three already completed, and the fourth currently underway. Already the DoE has praised the ever-lower bid prices, with the average energy price from bid window 1 to bid window 4 decreasing from R3 228/MWh to R786/MWh (adjusted for April 2014 base).
Because the lowest RE price per kWh at the point of connection to the grid is so important in the bid process, solar PV bidders generally opt for areas with the highest irradiation and thus the best energy yields – typically in the Solar Corridor. In the process, the bid criteria ignores the costs of additional grid infrastructure and upgrades, and transmission losses to the point of consumption. This results in grid congestion, time delays and importantly, additional real delivered costs when the transmission grid upgrades and transmission losses are included.
Part of the problem is that the IRP 2010-2030 and REIPPP programme do not spatially define the generation capacity allocations – and this is important not only in cost terms, but also for transmission grid planning, design and grid stability.
Fig. 3: Solar irradiation map of South Africa. [Credit: GIZ / SolarGIS]
Holistic, viable, spatial approaches
A GIZ study (German Collaboration) for the DoE and Eskom compared the current bidding criteria (and its resulting allocation of solar PV bids) with two alternative PV allocation strategies. In the GIZ study report, “Analysis of options for the future allocation of PV farms in South Africa” (published in March 2015), the authors considered the total levelised cost of electricity for the allocation of solar PV bids based on three options, namely:
• Scenario A: Current allocation strategy predominantly in the Solar Corridor (maximum energy yield);
• Scenario B: Allocation predominantly close to load centres (minimum grid reinforcements); and
• Scenario C: Allocation predominantly within defined Renewable Energy Development Zones (REDZs).
REDZs have been identified by the CSIR and Department of Environmental Affairs (DEA) as part of the Strategic Environmental Assessment project (see Fig. 4). The project integrates environmental, economic and social factors to identify geographical areas where in the medium to long-term wind and solar PV development will have the lowest impact on the environment while yielding the highest possible social and economic benefit.
The REDZs also take into account transmission losses, municipalities with high social need and development potential, priority areas for renewable energy manufacturing and import activities, and existing transmission infrastructure. The once-off environmental pre-assessment of designated areas further help reduce the necessity to repeat the process for each individual project, and solar PV and wind developers, provincial governments, the private sector and the public were consulted in the refinement of REDZs.
In Scenario A, 6 400MW is considered within the Solar Corridor, with the remaining 2 000MW distributed across the country. In Scenario B, 2 600MW is considered within the Solar Corridor, with the remaining 5 800MW distributed across the country close to major load centres. In Scenario C, 2 800MW is considered in the Solar Corridor, with the remaining 5 600MW distributed predominantly in the REDZs.
Fig. 4: Proposed REDZs [Credit: CSIR]
In the GIZ study, the total levelised cost of RE (US$/kWh) for the current and above two alternative allocation strategies were calculated, comprising the levelised cost of RE at the point of connection, plus the required transmission grid upgrades, plus transmission system losses based on load-flow simulations.
The GIZ study concludes that the total levelised costs of RE for all three solar PV bid allocation options, when considered holistically, are very similar. The authors thus argue that the other allocation options are not only viable, but would mean less grid congestion with timelier realisation of projects.
This means the geographic placement of solar PV for most parts of SA does not significantly or negatively impact on the total levelised cost of RE, and that other considerations such as the environmental and socio-economic aspects, can be better accommodated and taken into consideration.
Fig. 5: Total levelised cost of electricity of solar PV generation. [Credit: GIZ]
The current bid selection criteria unfairly favours projects with a low levelised cost of RE at the point of connection to the grid, while the alternative allocation strategies result in better distribution of solar PV allowing much-needed generation capacity to be connected faster, through minimisation of transmission grid upgrades and transmission losses.
While the GIZ study only considered solar PV, the same arguments can perhaps also be applied to other RE technologies, including concentrating solar power and wind power, where similar grid congestion, delays and transmission losses occur as a result of the best RE yield sites occurring in limited geographic areas far from load centres.
Rethinking the bid selection and allocation process
Both IRP 2010-2030 and its RE implementation vehicle, the REIPPP programme, are outdated and need serious rethinking. IRP 2010-2030 was needed to chart an energy generation course for South Africa, and the REIPPP programme for initiation and implementation of a major, utility-scale RE procurement. Had we waited for the perfect strategy, progress toward implementation would have been delayed even further.
The IRP 2010-2030 and its REIPPP programme counterpart are spatially blind, and yet the implementation of RE generation is a spatially-based process – not only in the siting of RE generation plant itself, but also in the geography of the transmission grid and the location of the load centres for the electricity generated.
The geographic blindness of the current bid criteria is exacerbating current generation capacity constraints. A flawed bid evaluation process results in bidders choosing the highest-yield locations with the lowest-cost of RE at the point of connection to the grid, which is not necessarily optimal for minimising grid congestion, connection delays, transmission grid upgrades and transmission losses.
The IRP 2010-2030 set out to be “living plan” that was meant to be updated every two years, but it is now five years old, without any DoE or cabinet approved update. The REIPPP programme bid criteria are also overdue for reconsideration. At the least a more holistic approach is needed for the selection and award of successful bids.
Spatial approaches provide ways in which the growing spatial extent of renewable energy development in can be managed sustainably and cost-effectively. It also provides a means to achieve a spread of development, both technically and socio-politically as outlined in the South African National Development Plan (NDP) and New Growth Path (NGP) for South Africa and National Infrastructure Plan (NIP).
The hard work has already been done and the numbers weighted. This makes it possible to speed up RE project implementation, do so efficiently, and at a better-calculated cost. Even more importantly, REDZs take into account a more holistic approach with regard to the economic, social and industrial development of South Africa, while taking the full cost chain into account.