3C Tecnología. Glosas de innovación aplicadas a la pyme. ISSN: 2254 – 4143 Ed. 39 Vol. 10 N.º 3 Septiembre - Diciembre 2021
39
INVESTIGATING ALTERNATIVE POWER GENERATION
STRATEGIES FOR LOCAL MUNICIPALITIES THAT ARE TIED TO
THE NATIONAL GRID
Bonolo Patricia Tshetlhe
Department of Electrical, Electronic and Computer Engineering,
Central University of Technology, Free State Private Bag X20539, Bloemfontein, (South Africa).
E-mail: tshetlhebt@gmail.com ORCID: https://orcid.org/0000-0002-8468-4694
Arthur James Swart
Central University of Technology, Free State, Bloemfontein, (South Africa).
E-mail: aswart@cut.ac.za ORCID: https://orcid.org/0000-0001-5906-2896
Phillip Koko
Central University of Technology, Free State, Bloemfontein, (South Africa).
E-mail: skoko@cut.ac.za ORCID: https://orcid.org/0000-0001-8954-0588
Recepción:
07/06/2021
Aceptación:
31/08/2021
Publicación:
14/09/2021
Citación sugerida:
Tshetlhe, B. P., Swart, A. J., y Koko, P. (2021). Investigating alternative power generation strategies for local
municipalities that are tied to the national grid. 3C Tecnología. Glosas de innovación aplicadas a la pyme, 10(3), 39-55.
https://doi.org/10.17993/3ctecno/2021.v10n3e39.39-55
3C Tecnología. Glosas de innovación aplicadas a la pyme. ISSN: 2254 – 4143 Ed. 39 Vol. 10 N.º 3 Septiembre - Diciembre 2021
40
https://doi.org/10.17993/3ctecno/2021.v10n3e39.39-55
ABSTRACT
Escalating electrical energy usage and costs over the past few years has resulted in large utility bill expenses
that municipalities are struggling to pay-o to National Energy Suppliers. Furthermore, some energy
suppliers are struggling to meet the demand for more energy due to a variety of factors. The challenge
therefore exists in identifying viable alternative power generation strategies for local municipalities to
reduce their current electrical energy expenses or to provide limited power to their community when
disconnected from the National Grid. An environmentally friendly renewable energy strategy method
could be used to supplement the current energy requirements of a municipality during months of high
energy demand. The main focus of this study will be on a small town in the Free State province of South
Africa, called Koefontein. A battery-based solar PV system was designed in the Homer software and
chosen as the renewable energy strategy to supplement the current energy needs of Koefontein due to
its performance and cost eectiveness. The initial implementation cost of the system is $ 42 995 649.95.
The cost of energy for the PV system suggested for Koefontein houses is $ 0.40/kWh with the yearly
electricity production of 27 661 kWh. The payback period of the system is 45.3 years. The municipality
needs to consider installing battery-based solar PV system to supply businesses in Koefontein during
their high demanding hours and during load shedding as the system indicates an aordable cost of
energy with high yearly production.
KEYWORDS
Solar PV Systems, Wind Turbine Systems, Load Shedding, Energy Audit.
3C Tecnología. Glosas de innovación aplicadas a la pyme. ISSN: 2254 – 4143 Ed. 39 Vol. 10 N.º 3 Septiembre - Diciembre 2021
41
https://doi.org/10.17993/3ctecno/2021.v10n3e39.39-55
1. INTRODUCTION
Electricity supply in South Africa (SA) has long been the domain of the National Energy Supplier, called
Eskom (Nehrir, 2011). Eskom supplies 96% of electricity in SA (Alfreds, 2018). Eskom and municipalities
both distribute electricity to consumers; the distribution function is shared between them (Eskom, 2018). It
is a major source of income for municipalities that supply electricity to households and businesses (South
African Government, 2011). Eskom supplies the licensed municipalities in bulk at a pre-determined
tari, then the municipality re-sells electricity to the end users within their municipal borders at a mark-
up. There are municipalities which are struggling to pay Eskom the amount of money owed due to low
revenue collection from electricity e.g. Matjabeng local Municipality is currently indebted to Eskom
for almost 2 Billion Rand ($ 105 654 156), part of which has been outstanding and in escalation since
October 2007 (Eskom, 2018). Energy autonomy is an option being seriously considered now more than
ever before and for good reasons.
The main aim of this study is to investigate alternative power generation strategies for local municipalities
in order to enable more autonomy and that can help to reduce the pressure placed on the National Gird.
This paper will discuss the importance of renewable energy resources. Data requirements for Homer
will be analyzed and simulation results of two main strategies (solar and wind) will be presented. The
conclusions end the discussion.
1.1. CONTEXT THAT NECESSITATES ALTERNATIVE STRATEGIES
The demand for power keeps growing at an alarming rate while supply trails behind. This leads to the
implementation of load shedding initiatives to keep the country illuminated. When there is not enough
electricity available to meet the demand for all Eskom customers, it could be necessary to interrupt
supply to certain areas; this is called load shedding (Davidson, 2014). Load shedding is dened as a
coordinated set of controls that decreases the electric load in one part of the system to restore the overall
system back to its normal operation conditions (Swart, 2018).
3C Tecnología. Glosas de innovación aplicadas a la pyme. ISSN: 2254 – 4143 Ed. 39 Vol. 10 N.º 3 Septiembre - Diciembre 2021
42
https://doi.org/10.17993/3ctecno/2021.v10n3e39.39-55
In November 2007, load shedding hit SA for the rst time, disrupting businesses, closing mining
operations and aecting households (Coetze & Mart-Mari, 2016). The national power grid again
came under severe constrains during the 2013/2014 summer maintenance program, requiring Eskom
to implement load shedding again. Eskom implemented 99 days of load shedding in 2015, causing a
decrease in manufacturing and mining output, dragging down economic growth (Coetze & Mart-Mari,
2016).
Renewable and clean alternative power generation technologies can play an important role in mitigating
these occurrences of load shedding. Increased global public awareness of the need for environmental
protection and desire for less dependence on fossil fuels for energy production is also required (Nehrir,
2011). SA has to consider alternative power generation strategies, such as solar and wind energies, to
keep up with the growing demand and to enable a better level of sustainability (Nehrir, 2011).
SA has one of the best solar irradiances in the world and experiences some of the highest levels of yearly
horizontal solar irradiation globally. The average daily solar radiation in SA is between 4.5 and 6.5
kWh/m2/day (Niselow, 2019). In terms of SA’s theoretical wind potential, research from the Council
for Scientic and Industrial Research suggest that to generate the equivalent of SA’s current electricity
demand, only 0.6% of the available SA’s land mass would have to be dedicated to wind farms (Nehrir,
2011). The two main alternative strategies for this study focus on the use of wind and solar farms as
possible supplements to the current energy needs of one town in the Letsemeng Local Municipality in
the Free State Province of SA.
2. METHOD
This research is focused on one town in the Letsemeng Local Municipality, located in the Free State
Province of SA, which is Koefontein. The objective of this research is to:
1. Obtain the energy usage bills from Eskom to determine what energy needs to be supplemented.
3C Tecnología. Glosas de innovación aplicadas a la pyme. ISSN: 2254 – 4143 Ed. 39 Vol. 10 N.º 3 Septiembre - Diciembre 2021
43
https://doi.org/10.17993/3ctecno/2021.v10n3e39.39-55
2. Determine the number of homes/businesses that contribute to this bill, to determine if both or
just one can be supplemented.
3. Recommend an appropriate renewable energy strategy (solar or wind) for the identied
municipality.
The energy bills from Eskom to the municipality are obtained at the main meter of Eskom in each
town. This bill indicates power consumption of the municipality for a certain month and the Notied
Maximum Demand. For winter months, the energy is classied into two groups: the low and high season
energy and they dier in charges.
An energy audit is conducted to determine the energy required by homes and businesses in the two
towns. Audit levels clearly dier in respect of their set of objectives, scope of tasks and powers and the
related tools (Palyi, 2015). A novel approach would involve using Google Maps as a tool to determine
the number of households and businesses that contribute to the present energy demand. This forms a
basic level audit.
Household neighborhoods are identied by swimming pools, sports grounds and schools. Businesses
would be identied by the nearby presence of government departments of public buildings. This is then
correlated to the amount of electrical energy that was sold to residential and industrial businesses for the
past year. This helps to identify months of high energy usage. This data is then correlated to the weather
data (solar radiance and wind speed per month) that has been obtained for the past 10 years in order to
determine if a solar or wind farm is more suitable. After comparing the annual weather data with the
monthly energy consumption from the municipality, an alternative power generation method can be
suggested. Supplementary energy could be provided during high demanding hours of businesses and
when government departments and schools are operative.
The Homer program can be used for simulating the output power of the two strategies with the
given weather data. This software has been used both to analyze the o-grid electrication issues in
3C Tecnología. Glosas de innovación aplicadas a la pyme. ISSN: 2254 – 4143 Ed. 39 Vol. 10 N.º 3 Septiembre - Diciembre 2021
44
https://doi.org/10.17993/3ctecno/2021.v10n3e39.39-55
the developed as well as developing countries (Sen, 2013). Solar radiation data, wind speed data and
electricity consumption of the municipality needs to be used as input data to Homer. A cost analysis and
payback period are then computed.
2.1. DATA REQUIREMENTS
During the audit, the following identication criteria was used: household neighborhoods were identied
by the presence of a school (sports ground or label indication). Businesses were identied by the presence
of a government building or label indication. The results of the audit indicated that Koefontein consist
of 2038 houses and 113 businesses, with the houses consuming more energy than businesses.
The results of the audit were correlated to the municipal electricity sales that are shown in Figure 1. June
2016 has the highest energy consumption for Koefontein houses, therefore the energy consumption
obtained for June 2016 is used to create a daily load prole for Homer. For the year 2016, Koefontein
had a total revenue collection of R 9 277 146 ($ 613 767), of which R 6 893 384 ($ 456 059) was
collected from houses and R 2 383 761 ($ 157 707) was collected from businesses.
Figure 1. Total energy consumption of Kofefontein town.
Source: own elaboration.
3C Tecnología. Glosas de innovación aplicadas a la pyme. ISSN: 2254 – 4143 Ed. 39 Vol. 10 N.º 3 Septiembre - Diciembre 2021
45
https://doi.org/10.17993/3ctecno/2021.v10n3e39.39-55
Energy consumption data obtained from the municipality is in months. To convert the energy consumption
to daily consumption it was rstly converted to weekly consumption using equation 1.
Weekly energy consumption =
(1)
=
=
The total daily energy consumption is then 40 556. 171 kWh (weekly divide 7) and the daily energy
consumption per house is then 19.9 kWh (total daily divide 2038). The daily usage of an average
household in Koefontein town is used to design a 24-hour electricity consumption required as input
data for Homer. The load peak demand for the average house in Koefontein is 2 kW. When summing
the hourly usage for 24 hours, as indicated in Figure 2, then a total daily energy consumption of 19.9
kWh is obtained. The peak load requirement decides the size, structure and architecture of the proposed
system (Belu, 2014). The size of the alternative energy system should consider the month of June 2016
which is the highest energy consuming month for Koefontein. Homer software is used to design a
system for 1 house in Koefontein town. The results can then be scaled to cover all 2038 houses in
Koefontein.
2.0
1.5
1.0
0.5
0.0
0 6 12 18 24
Figure 2. Hourly energy consumption of an average sized home in Kofefontein.
Source: own elaboration.
3C Tecnología. Glosas de innovación aplicadas a la pyme. ISSN: 2254 – 4143 Ed. 39 Vol. 10 N.º 3 Septiembre - Diciembre 2021
46
https://doi.org/10.17993/3ctecno/2021.v10n3e39.39-55
3. RESULTS
3.1. RESULTS FOR A SOLAR ENERGY SYSTEM
Solar radiation data is one of the important inputs that is required by Homer when doing simulations
for solar PV systems. Figure 3 indicates the solar radiation data that was obtained from the Pulida solar
plant located about 20 minutes’ drive outside Jacobsdal, a small rural village in the Free State Province
of South Africa. This data was collected in Jacobsdal town which is the nearest town to Koefontein
because there is no solar plant in Koefontein. The scaled annual average of the site is indicated to be
6.06 kWh/m²/d, this value is calculated automatically by Homer.
Daily Radiation
Clarness Index
Figure 3. Solar radiation graph from HOMER based on the data from Jacobsdal.
Source: own elaboration.
Figure 4 shows the proposed battery-based solar PV system to be used to supply an averaged sized
home in Koefontein. The PV panels have no tracking device and they were modelled with a slope of
3C Tecnología. Glosas de innovación aplicadas a la pyme. ISSN: 2254 – 4143 Ed. 39 Vol. 10 N.º 3 Septiembre - Diciembre 2021
47
https://doi.org/10.17993/3ctecno/2021.v10n3e39.39-55
30 degree. The following individual PV panel sizes were considered in integer steps of one kW from
1-20 kW. The price of a 1 kW PV panel was considered to be R 7 592 ($ 210 96). The operation and
maintenance (O&M) cost of the solar PV panels is assumed to be 1.56 % of the capital cost (Garni,
2017). 58 Ah Trojan T-105 battery was considered for this study. The individual sizes were considered in
integer steps of one Ah from 1-40 Ah. The price of the battery was taken to be R 2 894 ($ 189) with the
O&M cost assumed to be 2% of the capital cost of the battery (Koko, 2014).
A 5-kW inverter was considered to meet the peak demand of the load. The price of the pure sine wave
inverter which is a Bi-Directional designed to obtain optimum inverter AC power from an installed DC
battery system is found to be R 29 599 ($ 1 958). The O&M cost is assumed to be 1% of the capital
cost (Koko, 2014). The typical lifespan of the converter is considered to be 15 years (Koko, 2014). The
load demand of the studied averaged sized home is found to be 20 kWh/day, as obtained from the
municipality. During Homer simulation, the daily random variation of 10% was considered since it is
impossible for the daily load demand to be constant throughout the year. The system indicated in Figure
4 has an electricity production of 27,661 kWh/year as indicated in Figure 5.
Figure 4. Proposed solar PV system to supply 1 house in Kofefontein.
Source: own elaboration.
3C Tecnología. Glosas de innovación aplicadas a la pyme. ISSN: 2254 – 4143 Ed. 39 Vol. 10 N.º 3 Septiembre - Diciembre 2021
48
https://doi.org/10.17993/3ctecno/2021.v10n3e39.39-55
Figure 5. Yearly electricity production.
Source: own elaboration.
This system consists of a PV of size 16 kW, 58 Trojan T- 105 battery which are connected in 2 batteries
per string. Excess electricity for this system is 18,817 kWh/year. This is found to be 68% of the overall
generated PV electrical energy. Hence, only 32% of the generated PV electrical energy has been utilized.
Since this excess electricity is not utilized for the load demand, the municipality can sell it into the Eskom
utility grid to generate more revenue.
3.2. RESULTS FOR A WIND SYSTEM
For this study, the wind data was collected from one of the South African Weather Services station in
Fauresmith (Free State). This town is located 54.5 km away from Koefontein. The wind speed data was
collected at dierent times of the day, this data can help to suggest if a wind renewable energy system
3C Tecnología. Glosas de innovación aplicadas a la pyme. ISSN: 2254 – 4143 Ed. 39 Vol. 10 N.º 3 Septiembre - Diciembre 2021
49
https://doi.org/10.17993/3ctecno/2021.v10n3e39.39-55
would be able to supplement the current energy needs of the two towns. Figure 6 indicates the wind
speed data obtained at 20:00 pm. Data collected at 20:00 pm was chosen because when most of the
people are back from work and schools (between 17:00 pm and 21:00 pm), the peak load demand takes
place. The annual average wind speed is 1.53 m/s at an anemometer height of 10 m.
Figure 6. Wind speed data obtained at 20:00 pm from Fauresmith.
Source: own elaboration.
Figure 7 shows the proposed battery-based wind turbine system to be used to supply an averaged
sized home in Koefontein. This system indicates the primary load of 20 kWh/d which is used as the
electricity consumption input for an average house in Koefontein. During Homer simulation, wind
turbine system sizes were considered in steps of 10 kW integer values ranging from 0 to 23 wind turbines.
The capital cost of the generic 10 kW was assumed to be R 465 320 ($ 307 85) and the O&M of the
system was taken as 2% of the capital (Koko, 2014). The lifetime of the system was assumed to be 25
years. The following individual sizes were considered for the Trojan T-105 battery: steps of one Ah
integers from 3501-3520 Ah. The same capital cost of the battery used for the PV system was considered.
Figure 8 indicates the yearly electricity production results of the battery-based wind turbine system. This
system has an electricity production of 1 907 kWh/year. This has resulted in an excess electricity of 3.26
3C Tecnología. Glosas de innovación aplicadas a la pyme. ISSN: 2254 – 4143 Ed. 39 Vol. 10 N.º 3 Septiembre - Diciembre 2021
50
https://doi.org/10.17993/3ctecno/2021.v10n3e39.39-55
kWh/year (0.171%) with an unmet electricity load of 5.21 kWh/year. This system consists of 170 kW
wind turbine systems and 7012 Trojan T-105 batteries which are connected in 2 batteries per string.
Figure 7. Proposed wind turbine system to supply 1 house in Kofefontein.
Source: own elaboration.
Figure 8. Yearly electricity production of the wind turbine system.
Source: own elaboration.
3C Tecnología. Glosas de innovación aplicadas a la pyme. ISSN: 2254 – 4143 Ed. 39 Vol. 10 N.º 3 Septiembre - Diciembre 2021
51
https://doi.org/10.17993/3ctecno/2021.v10n3e39.39-55
Before deciding which system to install for houses in Koefontein, payback period of each system
suggested by Homer for Koefontein houses is determined. To determine the payback period of the
municipality, revenue collect from household electricity sales, total cost of the PV system which is R 649
883 554 ($ 42 995 649) for 2038 houses is used and total cost of the wind turbine system which is R 57
538 299 664 ($ 380 067 671 2) for 2038 houses is also used.
An averaged sized home in Koefontein requires 16 kW PV panels and 170 kW wind turbine system to
meet its load demand. To determine the size of the PV panels and wind turbine system required to meet
the load demand of the total number of houses which is 2038, total number of houses which is 2038 is
multiplied by the 16 kW PV panels and 170 kW wind turbine system respectfully, this gives a total of 32
608 kW PV panels and the wind turbine size is 346 460 kW. Equation 2 is used to calculate the payback
period of the municipality for both the battery-based wind turbine system and battery-based solar PV
system for 2038 houses. The payback period for the battery-based solar PV system is 45.3 years, this is
the maximum period which can be reduced if the municipality wishes to sell the excess energy produced
back to the national energy supplier in the country.
Cost of the system for the battery-based solar PV system is R 649 883 554 ($ 42 995 649) and for the
battery-based wind turbine system is R 57 538 299 664 ($ 380 667 671 2), these amounts are taken as
the initial investment respectfully. Revenue collected in 2016 which is R 6 893 384 ($ 456 059.49) and
revenue collected in 2017 which is R 7 437 498 ($ 492 057) is combined and taken as cash inow per
period.
Payback period =
Payback period (Solar PV) = = 45.3 years
Payback period (wind turbine) =
= 4014.10 years
3C Tecnología. Glosas de innovación aplicadas a la pyme. ISSN: 2254 – 4143 Ed. 39 Vol. 10 N.º 3 Septiembre - Diciembre 2021
52
https://doi.org/10.17993/3ctecno/2021.v10n3e39.39-55
4. CONCLUSIONS
The main aim of this study was to investigate alternative power generation for local municipalities in
order to enable more autonomy and that can help to reduce the pressure placed on the National Gird.
To meet the load demand of 2038 houses in Koefontein a battery-based solar PV system of size 32 608
kW is required, or a battery-based wind turbine of size 346 460 kW is required. The initial cost of the
battery-based solar PV system is R 649 883 554 ($ 42 995 649) while the initial cost of the battery-based
wind turbine system is R 57 538 299 664 ($ 380 667 671 2).
The cost of energy for the PV system suggested for Koefontein houses is R 6.10/kWh ($ 0.40) with
the yearly electricity production of 27 661 kWh. The wind turbine system has the cost of energy of R
776/kWh ($ 51.34) with the yearly electricity production of 1 907 kWh. The PV system has high excess
electricity of 18 817 kW while the wind turbine system has an excess electricity of 326 kW.
After performing the payback calculations, the battery-based solar PV system indicated to be the best
renewable energy system to be used in Koefontein. The cost of energy for the system is cost eective
for municipal customers. The municipality will collect more revenue after paying o the system. The
payback period maybe reduced if the municipality wishes to sell the excess energy produced back to the
national energy supplier in the country.
The municipality should consider doing an energy audit at least every two years to make sure that
they are aware of any energy loses so that it can be addressed before it aects revenue collection of
the municipality. The municipality should consider installing battery-based solar PV systems for the
businesses also so that they may have excess to electricity during load shedding and reduce the pressure
on placed on the national grid during high energy demanding hours of businesses.
This study does not include hydro power systems, biomass and geothermal renewable energies. This
research is only focused on the electricity sales of one municipality, it does not include electricity loses
3C Tecnología. Glosas de innovación aplicadas a la pyme. ISSN: 2254 – 4143 Ed. 39 Vol. 10 N.º 3 Septiembre - Diciembre 2021
53
https://doi.org/10.17993/3ctecno/2021.v10n3e39.39-55
due to tampered meters. Renewable energy alternatives such as solar and wind are gaining momentum
and will help in providing electricity to future new developments.
REFERENCES
Alfreds, D. (2018, October 29). Revealed: Here's where the worst air pollution is in SA. https://www.news2.
com/amp/news24/green/news/revealed-heres-where-the-worst-air-pollution-is-in-sa-20181029
Belu, D. R. (2014). Teaching renewable energy system designs and analysis with Homer. In 121st
ASEE Annual Conference and Exposition. Indianapolis. https://www.asee.org/le.server/papers/
attachment/le/0004/5010/ASEE-2014-Teaching_Renewable_Energy_System_Design_With_
HOMER.pdf
Coetze, D., & Mart-Mari, E. (2016). The impact of load shedding on the construction industry in
South Africa. In 9th cidb Postgraduate Conference. Cape Town, South Africa. https://openbooks.uct.
ac.za/uct/catalog/download/cidb/7/210-2?inline=1
Davidson, D. O. (2010). Recomendation to the World Bank Group on leading to South Africa for
Eskom Investmentsupport project that include a large coal burning power station at Medupi.
External peer review report on Eskom, 1-6. https://www.climateinvestmentfunds.org/sites/default/
les/expternal%20peer%20review%20report%20on%20Eskom.pdf
Eskom. (2018, February 8). Eskom holdings SOC ltd. https://www.eskom.co.za
Garni, H. Z. (2017). Optimal design and analysis of grid-connected photovoltaic under dierent tracking
systems using Homer. Energy convertion and management, 155(1), 1-57. https://dx.doi.org/10.1016/j.
enconman.2017.10.090
Koko, S. (2014). Micro-hydrokinetic for remote rural electrication. International journal of electrical,
Robotics, Electronics and Communication Engineering, 8(11), 1384-1388. https://doi.org/10.1016/j.
epsr.2015.04.018
3C Tecnología. Glosas de innovación aplicadas a la pyme. ISSN: 2254 – 4143 Ed. 39 Vol. 10 N.º 3 Septiembre - Diciembre 2021
54
https://doi.org/10.17993/3ctecno/2021.v10n3e39.39-55
Nehrir, N. (2011). A review of hybrid Renewable/Alternative energy systems of Electric Power
Generation Conguration,Control and Application. IEEE Transactions on sustainable energy 2(4),
393-399. http://dx.doi.org/10.1109/TSTE.2011.2157540
Niselow, T. (2019, March 24). Fin24. Load shedding through the years and how Eskom has struggled to keep the
lights on. https://m.n24.com
Palyi, A. (2015). The contribution of the state audit oce to good governance and renewable of
accounting. Public Finance Quarterly, State audit oce of Hungary, 60(4), 1. https://www.asz.hu/
storage/les/les/public-nance-quarterly-articles/2015/a_palyik_2015_4.pdf
Sen, R. (2013). O-grid Electricity generation with renewable energy technology in India: An application
of Homer. Renewable energy, 62(3), 1-15. http://dx.doi.org/10.1016/j.renene.2013.07.028
South African Government. (2011). Local government budgets and expenditure review. http://www.gov.
za/2011-local-government-budgets-and-expenditure-review
Swart, A. J. (2018). Final pass rate uctuations of power engineering students may be related to load
shedding and not load shifting. International Conference on Multidisciplinary Research, Mauritius. https://
www.researchgate.net/publication/326380796_Final_Pass_Rate_Fluctuations_of_Power_
Engineering_Students_May_Be_Related_to_Load_Shedding_and_not_Load_Shifting
