Monday, June 20, 2016

Diesel vs. Solar Generated Electricity in Botswana

Dumela*. As I have been travelling around Botswana visiting solar installations and chatting to solar system owners, I have encountered a lot of solar photovoltaic (PV) systems that have a diesel generator as back-up and so I have been fielding—and asking—a lot of questions about diesel versus solar. Most of my research has focused on solar systems with battery storage, but I decided that it was worth learning a little more about diesel generators: when and how they might be best combined with solar systems. In this blog, I share with you some of what I have learned.

Diesel (or petrol) generators have been around for a long time. They are robust and reliable, and there are numerous units from various manufacturers available in different sizes for different applications. They vary from small emergency generator units, which have a capacity of 2 to 5 kW, that you can purchase in a large supermarket along with your building supplies or groceries, to large units that serve as a back-up electricity supply for lodges, to even larger units that can power mining operations.

Diesel generators can even be used to add on-demand generating capacity to the electrical grid. For example, the Botswana Power Corporation (BPC)-owned 90 MW back-up generator operation in Orapa consists of two 45 MW GE LM 6000 turbine/generator units. Based on their specification data, when they are both running, these units can consume up to 22 000 liters of diesel per hour – that is almost one diesel tanker truck of fuel every sixty minutes! With diesel presently priced at about P 7.30/liter (US$2.52/gal), we can calculate that the Orapa station is, at capacity and just in terms of its fuel costs, generating electricity for ~P 1.78/kWh.

When considering off-grid electricity systems, the decision is usually made between selecting solar or a generator. In many respects, a diesel generator seems like a straightforward solution. The upfront costs are lower than those of a solar system; a generator is not dependent on sunlight and weather conditions; it does not need battery storage, and it is fairly easy to find technical assistance when the generator is not working. A small generator can even be loaded onto the back of a truck and transported to the repair shop.

On the down side, diesel generators are noisy, low-efficiency, highly polluting devices that require the transportation and storage of diesel. Typically, only about 25 to 37% of the energy in diesel is converted to electricity: the rest is lost as heat from the engine block and the hot exhaust gases.

Another complicating aspect of diesel generator operation is that generators burn a minimum amount of fuel, even at very low electrical demands. As Figure A shows, for a range of generator sizes, fuel consumption increases as the electrical load increases; however, at very low electricity demand, the consumption levels off. As a result, generators can be horribly inefficient (<15%) and wasteful devices at low loads. This escalates the cost of producing electricity at low demand levels, as shown in Figure B. If you need to turn on just a single light, you will need to fire up the generator–which will consume a minimum amount of fuel. It is for this reason that generators are often paired with some battery storage that can take care of minimum load requirements. Also bear in mind that these cost data were calculated using a relatively low diesel cost of P 7.30/liter: as recently as 2014, diesel was selling for over P 10/liter (see Figure C below).


Data Source: Able Sales
Data: World Bank and OANDA

Figure B indicates that the fuel costs of generating electricity from diesel generators is of the order of P 2.00 to P 2.20/kWh at demand levels above 50%, but this is only part of the story. To get a sense of the overall true costs of diesel generation, we need to include the purchase and maintenance costs of the generator.

To calculate the all-in lifecycle costs, we need to calculate the levelized cost of energy or LCOE. This parameter takes into account the initial cost of the unit, its electrical output, the annual cost of fuel, annual maintenance, and an interest cost (normally set by the minimum rate of return one would like to earn on a project and is often equivalent to the bank loan interest rate to purchase the equipment). These costs are then distributed evenly over the life of the project to yield a single “levelized” cost for energy. The attraction of the LCOE approach is that it allows a side-by-side comparison of different projects and energy sources that have very different financial requirements and expense flows. For example, it allows comparison of a solar system with high capital but low operating cost with a diesel generator project that has low capital but high operating costs. LCOE provides a means for determining which project is better from a lifecycle energy cost point of view.

Calculating the LCOE from scratch can be challenging as it is a rather involved financial calculation, but there is a great site, from the National Renewable Energy Laboratory in the US,  that can do the calculations for you. I used this to calculate the LCOE for a diesel generator. My set of assumptions included the following:

Capital cost: P 5000/kW ($ 450/kW @ P 11/US$)
Capacity factor: 90% (Generator runs most of the time)
Project lifetime: 10 years
Interest rates: 10% (Prime (6%) + 4%)
Diesel consumption rate: 0.27 L/kWh
Costs for diesel: P 7.30/L ($ 2.41/gal, June 2016)
Cost per kWh: P 1.71/kWh ($ 0.17/kWh)
Annual maintenance costs: P 330/kW ($ 30/kW)



With these assumptions, the calculator yielded the results in the figure above, indicating that the LCOE for a diesel generator is of the order of $ 0.183/kWh or P 2..01/kWh. I then computed the variability of the LCOE with high and low levels of some of the assumptions and generated the chart below. 

 
The data indicate that the cost of energy from diesel is most sensitive to the cost of diesel. Large increases in diesel prices lead to much higher costs for generated electricity; big increases in the cost of the generator do not significantly impact the electricity cost. The chart shows, within this span of assumptions, that the range of LCOE for diesel generators varies from P 1.86 to P 3.38/kWh, with a median value of P 2.62/kWh. To put these values into context, it should be noted that the average resident in Botswana, using more than 200 kWh/month, is paying P 0.88 per kWh (US$ 0.08/kWh @ P 11/US$) for electricity from the grid. In other words, within our chosen range of assumptions, the median cost of diesel-generated electricity is three times that of electricity purchased directly from BPC. So, if you have a grid connection and you install a diesel generator, do not consider the generator as a substitute for BPC: it is there to be used only when electricity is not available during load shedding.

Returning to the central theme of this blog (that of solar vs. diesel), let’s have a look at the LCOE for an off-grid solar installation with battery storage. The assumptions that I used for this calculation are as follows:

Capital cost: P 60 050/kW ($ 5500/kW @ P 11/US$)
Capacity factor: 90% (With battery storage, the system operates most of the time, like a diesel system)
Project lifetime: 10 years
Interest rate: 10%
Cost for fuel: Free!
Annual maintenance costs: P 55/kW ($ 5/kW)

I used local prices for the installed cost of a solar system with battery storage that I obtained from vendors here in Botswana. I also chose 10 years as a project lifetime to allow a direct comparison with the diesel project and because this is the typical lifetime of batteries under ideal conditions.  However, I will note that battery lifetime in Botswana can be a lot shorter as a consequence of the impact of high ambient temperatures.

Using the above assumptions, the LCOE for solar power calculates at P 1.25/kWh. This is still higher than the cost of grid-supplied electricity from BPC, but is almost half the cost of diesel-generated electricity.  The chart below shows how the LCOE changes with capital costs: as expected the costs for solar generated electricity are highly sensitive to the upfront costs.



LCOE comparisons for diesel vs. solar with battery storage are summarized in the table below.

It should be noted that I have included some pretty significant assumptions in my calculations, but the key points to note from these data are that (i) the capital cost of solar plus battery storage is 12 times that of a diesel generator with the equivalent capacity; (ii) the lifetime costs of electricity for diesel generators are twice that of solar. These options are both more expensive than electricity from the grid in Botswana.

When considering diesel vs. solar, the decision comes down to one of high upfront costs and low long-term operating costs compared with low upfront costs and high operating costs. It is clear from the financial data that solar, even with the costs of battery storage included, is the better option. The biggest obstacle to investing in solar is finding that substantially larger amount of money for the one-time installation of the solar and battery storage system.

However, (as these matters often are) it is sometimes not a black or white decision. Sometimes a hybrid solution, using a combination of solar and diesel, is the correct one. I noted in an earlier blog the disappointment many purchasers of solar have experienced due to poor battery and system performance. As I read and learn more, I am becoming convinced that the correct solution for off-grid systems is not diesel or PV: it is a hybrid of both. A long run of overcast days can really compromise the electricity output of a PV/battery storage system, so installing diesel generators alongside solar systems in a hybrid configuration can compensate for these weather shortcomings.

There are several other reasons why hybrid systems are a good choice. These include:

  • Improved reliability – crucial for service-oriented businesses, such as tourist lodges, which cannot afford to be without electricity;
  • Reduced capital costs – some diesel-generation capacity reduces the size and cost of the battery bank and solar array;
  • Better for short-term high-demand applications, such as stoves, welding machines, hairdryers, etc.;
  • Reduced emissions compared with a generator-only solution;
  • Reduced exposure to future diesel price increases;
  • Modular solutions are possible. As PV and battery prices decrease, more modules and/or batteries can be installed, leading to reduced fossil fuel use;
  • More system design options become available, for example, PV and no storage with generator or PV plus battery storage and a generator;
  • A hybrid solution is a partially “green” solution, but still allows optimization of capital expenditure. It avoids overspending on capital costs or underspending and then suffering from very high fuel costs.

Hybrid systems combining solar and diesel generators are growing in importance.  A number of island communities use them and there are a growing number of mining companies powered by diesel generators and using millions of liters of diesel fuel per year to power their operations that are finding big savings by introducing some solar generation into their energy mix. If you are interested in exploring hybrid options, I encourage you to take to look at the Homer Energy website and their Homer modelling tool. This enables you to investigate a range of hybrid options, all the way from 100% diesel to 100% solar and calculate the LCOE for each option. In this way, you can determine the optimal combination of solar and diesel for your application.


In conclusion, hybrid options seem to offer the best of both worlds. A hybrid system should have upfront capital costs somewhere between those of a diesel-only solution and a solar/battery storage system; the LCOE should similarly lie between those of both options. In turn, the resulting system has a higher degree of reliability, less weather sensitivity, and permits the installation of more solar and battery storage as time and finances allow. However, a combined system will no doubt be more complicated to operate, so if you have any direct experience with hybrid systems and their performance in the field, please share it with us in the comment box below.

In the meantime, remember to turn off the lights when you leave the room. 

Tsamayang Sentle**
Mike Mooiman
mooimanm@franklinpierce.edu


(*Greetings in Setswana)
(**Go well or Goodbye in Setswana)