Getting a handle on the cost of solar PV

James Green, Solar Water Heating Chair

It should be simple in the highly competitive growing market of solar rooftop PV to get a handle on the costs but the reality is that despite budgeting costs being straightforward, due to regulatory confusion, restrictions on feed in tariffs (FIT’s), and amounts (kW) that can be fed in, as well as other factors, determining the return on investment (ROI), and potential payback (time to recover the investment), is complex. 

Large Scale PV Installations

Large utility scale PV (say 50MW and above) for power generation into the grid under an REIPPP contract is theoretically straightforward. You (at a simple level) cost the array of solar panels, the racking, inverters, and step up transformers, into high voltage, add in the costs of the land, the EIA, financing costs, labour and earthworks, all the legal and compliance aspects, and come up with a cost per kWh you can sell to the government over 20 years, based on the expected yield in turn determined by solar insolation and location. 

As has been seen the price per kWh to be paid (by Eskom) for each solar PV kWh has progressively fallen from Round 1 to Round 4.  This is attributable to both falling prices of materials and increasing competition.  For the investors and project managers, the ROI and IRR calculations are based on the variables of cost, electricity yield, and the price paid by Eskom over the period.

Very attractive ROI’s existed for Round 1, and have progressively fallen as prices to be paid per kWh have declined.

Small Rooftop Solar PV

At the opposite end of the spectrum, say a 3kW-6kW solar PV rooftop installation, where a home or small business is looking to save money and carbon emissions, and to use the solar PV power for their own consumption, and not feed back into the grid, it is again relatively straightforward to budget and calculate the ROI.  

Not surprisingly the costs per Watt are higher than large scale, as a guide at late 2015 approx. R18+/- per Watt installed, against R12+/- per Watt (or less) installed for utility scale, but the ROI is a more complex calculation than the utility scale. 

One has to factor in the weekends and public holidays (118+/-) days per annum when the power might not be used at all, and the option of battery storage to provide power at night or when the grid goes down (load shedding). Depending on the amount of power required, when the PV is not generating electricity, the battery storage can double the overall capex costs.

Despite the capex costs per Watt being higher than large utility scale, the costs per kWh saved are also far higher than the rate the REIPPP will receive.  With grid supplied power per kWh costs ranging from lows of R1,50 to highs of R 2,25 at 2015, and progressively increasing, the ROI needs to factor in the actual kWh costs saved, including reductions in environmental levies, service delivery charges, etc and VAT. 

Assuming that the end user of a small rooftop installation can actually consume the PV solar power generated, an approximate guide for payback is 5-7 years without battery back up.

Solar PV, Rooftop and Other, Above 6 kW to Utility Scale

In Europe a number of countries have encouraged rooftop solar PV through incentives (tax credits), subsidies, and most significantly through Feed in Tariffs (FIT’s). In Germany for example over 40GW has been installed in under 10 years, where the sun shines almost as little as the UK, and is approximately 50% of the sunshine in South Africa. 

South Africa in contrast to Europe faces a number of conundrums. Although solar radiation is arguably one of the best in the world for solar PV, the service delivery model from electricity is grossly distorted through municipalities cross subsidizing other services through reselling electricity supplied by Eskom. Buying power from Eskom on a Megaflex tariff at say R0,78 to R0,95 (depending on how you calculate it), and reselling to the public, both domestic and business at as high as R2,25 is a great earner, which the municipalities are reluctant to give up.  

The reality would appear to be that National Government through Nersa, DoE and Eskom are prepared to encourage REIPPP generation, but are adverse to the consumer taking power into their own hands, in the same way as local municipalities are reluctant to embrace rooftop solar PV FIT’s.

This, perhaps more than any other is the largest constraining factor on the potential huge growth of rooftop solar PV in South Africa, but in all probability it will happen anyway, as a result of the increases in prices of electricity. 

The Health Warnings for Mid Sized Rooftop Solar PV

As the rooftop solar PV market gathers momentum, partially as a result of load shedding, and as more and more new salespeople enter the market sensing the opportunity for a ‘quick buck’, the knowledge required by the potential investor to ensure that they do not get hoodwinked into buying a PV system that does not provide them with the ROI they hope for is essential. 

The problem of course is that the investor doesn’t have the time to become an expert, and is therefore liable to be persuaded to purchase a PV system that may be inappropriate for their needs.  The complaints that the SESSA Ombudsman is receiving are mainly on PV now rather than solar water heaters.

Solar Water Heaters along with other energy efficient measures, lighting, heating and cooling, are as a result of their lower capital cost and faster ROI, the first step before embarking on rooftop solar PV. (See last weeks blog ‘Confused by the cost of solar water heaters – what does it really mean?’). 

For the potential investor in rooftop solar PV, there are some essential guidelines to take into account for determining cost and the ROI.

  • Understanding electricity load profiles, time of use and peak
    • Analyzing the electricity bill gives the first required information, for overall consumption and costs per kWh. Large users will be considerably more complex than smaller, with different rates paid for peak, seasons, weekends, service delivery etc. 
    • If Eskom or the municipality can provide a load profile over the last year on a minute-by-minute basis, this information can be fed into a computer model to provide accurate analysis of the time of use and peak consumption.
    • If the information cannot be provided it is essential to undertake a monitoring exercise to analyze the load profiles.
    • Next consider whether any peaks outside of daylight generating hours can be shifted from those times to daylight hours. 
    • Remove heavy electricity use items from the equation where possible, for example electric boilers to solar thermal or heat pumps or a combination. 

Figure 1: Illustrative load profile with morning and evening peaks

  • Sizing the Solar PV to Maximize the ROI
    • Choose the size of the PV system in kW and compare against the load profiles. 
    • If the PV system is based on a peak consumption it may result in surplus power being generated from the PV system that cannot be used and cannot be fed back into the grid 
    • If battery storage is to be used, sizing the amount of storage required is essential to avoid overspend, or not having enough power when Eskom is unavailable.

Figure 2: Same kWh loads with PV system sized to peaks not taking into account time of use, and potential wasted PV kWh output

Figure 3: By rebalancing load profiles (before reducing kWh load through SWH, lighting, heating and cooling) and choosing a smaller PV system with greater kWh savings a better ROI than the larger system.

  • Understanding the Feed In Tariffs and Additional Expenditures
    • If FITs are available the rate per kWh fed back into the grid, needs to be computed including sizing the system to not only meet the end user kWh requirements, but also the minimums required to be eligible for FITs, which may result in the overall PV system being up to 180% of own use consumption.
    • Additional expenditure may be incurred in the acquisition of the existing transformers from Eskom, the liability associated with older transformers requiring replacement at some time in the future, and the increased costs in transforming from low voltage to high voltage lines. 
  • Do the Math
    • The budgeted cost of installed system, needs to be compared with electricity tariffs, current and projected at a realistic rate (Eskom price increases of 8% p.a.), to arrive at the payback point and Return on Investment over 10,20, 25 years.
    • Ongoing monitoring is needed using the various software packages to compare expected with the reality. 

Please note this list is not exhaustive but minimum requirements that should be looked for.


With solar water heaters the return on investment calculation, with a cost per kWh saved, or cost per litre of useable hot water required is relatively straightforward. As a general rule, bigger is better as long as the storage is matched to the kWh output of the solar collector, and the installed unit does actually produce hot water that can be used in winter.

With rooftop solar PV bigger may not actually be better, as the energy profiles will dictate what is the optimum size (not peak). The investment in solar PV is likely to be considerably larger than SWH, and the payback period much longer. The ROI ignoring the initial capital investment, is likely to mirror the ROI on the SWH but be lower.

Additional considerations on rooftop solar PV include downtime in businesses, and consequential losses that may arise from load shedding, where a cost benefit to the overall ROI needs to be built into the equation. 

Unfortunately in South Africa the rooftop solar PV model has to contest with a confused regulatory environment, where both Eskom and municipalities really have little interest in consumer renewables that are in conflict with their business models of selling power.  

If power saving is to occur it is on their terms, not on the consumers who are taking the initiative.  Over time as the price of electricity prices increase (inevitable), the Eskom dinosaur model will, whether they like it or not be forced to change, and service delivery models will have to find a different way of generating revenue. Let’s hope that clean renewable energy is not taxed.