You are on page 1of 89

Report for

Doc No: Rev Page 1 of 88


Contents
Figures 4
Tables 5
Definitions 6
ABBREVIATIONS 8
ABOUT THIS GUIDE 10
About the team 13
Stakeholders and roles 15
EXECUTIVE SUMMARY 16
1 Introduction 18
1.1 SSEG integration status in South Africa 18
1.1.1 South African electricity mix 18
1.1.2 SSEG uptake statistics and challenges in South Africa 18
1.1.3 Consideration of the impacts of COVID-19 relating to SSEG uptake 24
1.1.4 Existing practices and policies around SSEG integration 24
Identified gaps in municipal readiness – overview 29
2 Identified gaps in municipal readiness for SSEG integration 30
2.1 Shortage of personnel skills and capacity 30
2.2 The absence of appropriate tools to perform detailed grid impact studies 31
2.3 Limitations of the NRS 097-2-3 Simplified Connection Criteria 31
2.4 Existing specifications and technical requirements 31
2.5 Desired goals for municipal SSEG integration 32
Guidelines and capacity building – overview 33
3 Guidelines and capacity building 35
3.1 NRS 097 35
3.1.1 NRS 097-2 36
3.1.2 The NRS 097-2-1: Utility interface 37
3.1.3 The NRS 097-2-3: Simplified utility connection criteria for LV-connected generators 50
3.2 MPE guideline 54
3.3 Municipal training on simplified and detailed connection criteria 56
3.3.1 Simplified connection criteria: NRS 097-2-3 and MPE guideline 56
3.3.2 Detailed connection criteria: Detailed grid impact studies 57
CSIR Simplified load flow tool – overview 58
4 CSIR Simplified load flow assessment tool 59
4.1 Tool description 59
4.2 Tool validation 62
4.2.1 Configuration testing 62
4.2.2 Municipal test cases 63
4.3 Key outcomes and learnings 65
4.4 Recommendations and future development 66
CSIR hosting capacity analysis tool – overview 67
5 CSIR hosting capacity analysis tool 68

Report for
Doc No: Rev Page 2 of 88
5.1.1 Description of the developed Deterministic Hosting Capacity Tool 69
5.1.2 Description of the developed Stochastic Hosting Capacity Tool 70
5.2 Tool comparison of developed methods 72
5.2.1 Hosting capacity output results for two terminals 72
5.2.2 Computational time comparison 73
5.2.3 Analysis of results 74
5.3 Recommendations 75
Recommendations on NRS 097 – overview 76
6 Recommendations on NRS-097 78
6.1 Objective 78
6.2 The 350 kVA Limitation for Simplified Connection Criteria 78
6.2.1 Basis of concern regarding the 350 KVA limit 78
6.2.2 Basis of the capacity limit of 75% of transformer rating 79
6.2.3 Recommended adjustment to 350 kVA limit requirement 80
6.3 Classification of UPS Systems 81
6.3.1 The Problem of SSEG with UPS Systems 81
6.3.2 Recommended classification inclusion to the NRS documentation 83
6.3.3 Recommended SANS standards to consult regarding UPS systems 83
7 Resources 85
8 References 87

Report for
Doc No: Rev Page 3 of 88
FIGURES
Figure 1: Technology sector clusters of the CSIR 12
Figure 2: Organisational structure of focal cluster, centre, and research group 12
Figure 3: South Africa's electricity mix 2020 [6] 18
Figure 4: National uptake of SSEG in municipalities South Africa [8] 19
Figure 5: Provincial SSEG uptake summary [8] 21
Figure 6: Estimated capacity (MW) of registered SSEG system in South Africa [8] 22
Figure 7: Main challenges in SSEG integration into municipal distribution [8] 23
Figure 8: Registration and approval process for SSEG 26
Figure 9: Standard conditions with which NERSA-registered facilities must comply 27
Figure 10: Power curtailment during over-frequency [24] 44
Figure 11: Single meter installation [24] 49
Figure 12: Multiple meter installation separate metering (left) and separate embedded metering (right)
[24] 49
Figure 13: Summary of simplified connection criteria [20] 51
Figure 14: Flow chart of simplified connection technical evaluation criteria [20] 53
Figure 15: Simplified analysis method 55
Figure 16: Determining possibility of connection of LV generator 56
Figure 17: Homepage of CSIR simplified load flow tool 58
Figure 18: Load flow process 59
Figure 19: Load flow method 61
Figure 20: Single line diagram of example of radial three bus network configuration in load flow tool (Two
parallel distribution lines between bus 1-2 and two parallel transformers between bus 2-3) 62
Figure 21: Test case 1: 900 kVA SSEG application (Municipal test case 1 − Three bus system) 64
Figure 22: Test case 2: 560 kVA SSEG application (Municipal test case 2 – Four bus system) 64
Figure 23: Typical flow chart to calculate hosting capacity 69
Figure 24: Flow chart for the Deterministic Hosting Capacity Method 70
Figure 25: Stochastic hosting capacity method flow chart 71
Figure 26: Hosting capacity calculated on two different terminals using both the deterministic and
stochastic method 73
Figure 27: Computational time deterministic method vs stochastic method 74
Figure 28: Original classification of scheme for residential SSEG 82
Figure 29: Proposed classification for residential SSEG 83

Report for
Doc No: Rev Page 4 of 88
TABLES
Table 1: Licensing and registration details for various sizes of embedded generation systems 25
Table 2: Future goals in terms of the metrics used to determine municipal readiness for SSEG integration
32
Table 3: Voltage requirements for grid-connected SSEG 40
Table 4: Frequency requirements for grid-connected SSEG 41
Table 5: Harmonic distortion requirements for grid-connected SSEG 41
Table 6: Flicker requirements for grid-connected SSEG 42
Table 7: Power factor requirements for grid-connected SSEG 42
Table 8: Information of reference source impedance, fault levels, and current injection requirements for
grid-connected SSEG 43
Table 9: Maximum harmonic current distortion as percentage of rated current (from NRS 097-2-1) 43
Table 10: Response to abnormal voltage and frequency conditions for SSEG in sub-categories A1 and
A2 [19] 48
Table 11: Comparison of single and multiple meter installations 50
Table 12: Recommendations on when to conduct detailed interconnection studies 57
Table 13: Tool properties 60
Table 14: Percentage similarity between load flow tool and DigSILENT PowerFactory 63
Table 15: Results for municipal test cases 65
Table 16: Comparison of deterministic and stochastic hosting capacity methods 67
Table 17: Comparison of the deterministic and stochastic hosting capacity methods 72
Table 18: Resources that provide SSEG integration information 85

Report for
Doc No: Rev Page 5 of 88
DEFINITIONS
A termination criterion chosen where the difference between the successive
Convergence
iterations in a load flow calculation have a difference in value smaller than or
criterion
equal to a chosen value, i.e. the values appear to be quasi-steady.
Convergent A steady load flow solution in terms of unknown voltages and power flows in
solution which the convergence criterion is satisfied.
In the context of this information booklet, this refers to customers who
Customer
consume electricity from the grid and generate electricity from their SSEG.
Dedicated
Feeder in a network that supplies a single customer or generator [1].
feeder
Distribution grid Network infrastructure operating at less than or equal to 132 kV.
Distribution A licensee authorised to operate and maintain a distribution system for
licensee supplying electricity to the consumers in the approved area of supply.
One or more energy generation sources that include the energy conversion
Embedded device(s), the static power converter(s), if applicable, and the control and
generator protection gear within a customer’s network that operate in synchronism with
the utility’s supply.
Maximum change in active power flow at the point of utility connection for a
Generator size generator trip (or rapid reduction in output) when generating at full active
power output.
Grid impact A study of impact in capacity and load effects of the main utility grid when
study connected to a source of renewable energy [2].
Low voltage Nominal voltage up to and including 1 kV.
Medium voltage Nominal voltage levels greater than 1 kV and less than 44 kV.
A distribution network that has one power source for a group of customers,
Radial network
and in the event of a power failure the entire group loses power.
Service Low voltage feeder (often single phase) from three-phase metering kiosk up
connection to the customer’s point of connection.
feeder
Low voltage feeder that supplies more than one customer and starts at the
Shared feeder medium voltage/low voltage transformer extending up to all three-phase
metering kiosks.
Embedded generator rated at up to 1 MVA which includes the energy
Small-scale
conversion device(s), the static power converter(s), if applicable, and the
embedded
control and protection gear within a customer’s network that operates in
generator
synchronism with low voltage networks.

Report for
Doc No: Rev Page 6 of 88
Small-scale An electricity payment structure that charges customers for energy purchases
embedded that use the national grid and compensates customers for the excess
generator tariff electricity fed by them onto the grid from their SSEG system [3].
The interconnected network of cables, transformers and equipment
Utility network associated with power transmission and distribution covering all voltage
ranges supplying consumers with electricity [4].

Report for
Doc No: Rev Page 7 of 88
ABBREVIATIONS
ADMD After Diversity Maximum Demand
DC Direct current
DPL DigSILENT Programming Language
EMC Electromagnetic compatibility
ERA Electricity Regulations Act
ESI Electricity supply industry
ESS Energy storage systems
HV High voltage
IRP Integrated Resource Plan
LV Low voltage
MPE Moeller & Poeller Engineering
MV Medium voltage
NERSA National Energy Regulator of South Africa
NMD Notified maximum demand
NRS National Rationalised Specifications
POC Point of connection
PV Photovoltaic
RPP Renewable Power Plants
SALGA South African Local Government Association
SANS South African National Standard
SSEG Small Scale Embedded Generation
UPS Uninterrupted Power Supply

Report for
Doc No: Rev Page 8 of 88
Report for
Doc No: Rev Page 9 of 88
ABOUT THIS GUIDE

The purpose of this information booklet is to provide information on the integration of Small-Scale
Embedded Generation (SSEG) in South Africa for municipalities, especially those experiencing an
increase in customers who wish to install SSEG onto municipal networks. The way the document is
organised is modular; following the introduction, each main section is preceded by an overview and
then more details follow the overview for the interested reader.
The topics covered in this document include the status of SSEG integration in South Africa in terms
of the current electricity mix, relevant statistics, and challenges in SSEG uptake, existing practices,
and policies around SSEG integration such as the NRS-097 and the Moeller & Poeller Engineering
(MPE) guideline. Gaps in municipal readiness to integrate SSEG are highlighted, given a gap
analysis performed on select municipalities as well as desired goals in terms of readiness for SSEG
integration. Established guidelines that assist in SSEG integration are discussed and municipal
capacity building activities performed by the CSIR are mentioned. CSIR tools are presented that
attempt to assist municipalities where a lack of tools for grid impacts is evident. Recommendations
are made on the NRS 097 series as well as resources section in the document for further information
on SSEG integration.

Report for
Doc No: Rev Page 10 of 88
ABOUT THE CSIR
“The Council for Scientific and Industrial Research (CSIR) is a leading scientific and technology
research organisation that researches, develops, localises and diffuses technologies to accelerate
socioeconomic prosperity in South Africa. The organisation’s work contributes to industrial
development and supports a capable state.”[5]
The CSIR was established in 1945 through an Act of Parliament. The executive authority of the
organisation is the Minister of Higher Education, Science and Technology. Key roles played by the
organisation include supporting the public and private sectors through directed research aligned with
South Africa’s priorities, the CSIR’s mandate and its competencies in science, engineering, and
technology.
The CSIR’s mandate [5]:
“The objects of the CSIR are, through directed and particularly multi-disciplinary research and
technological innovation, to foster, in the national interest and in fields which in its opinion should
receive preference, industrial and scientific development, either by itself or in co-operation with
principals from the private or public sectors, and thereby to contribute to the improvement of the
quality of life of the people of the Republic, and to perform any other functions that may be
assigned to the CSIR by or under this Act.” (Scientific Research Council Act 46 of 1988, amended
by Act 71 of 1990)
The core values of the CSIR are Excellence, People-centered, Integrity and Collaboration (EPIC
values)
“At the core of business of the CSIR is impact and the following objectives ensure that the CSIR’s
mission is achieved:

• Conduct research, development and innovation of transformative technologies and


accelerate their diffusion.
• Improve the competitiveness of high-impact industries to support South Africa’s re-
industrialisation by collaboratively developing, localising, and implementing technology.
• Drive socioeconomic transformation through research, development and innovation that
supports the development of a capable state.
• Build and transform human capital and infrastructure.
• Diversify income, and maintain financial sustainability and good governance.”

The technology sector clusters of the CSIR are shown in Figure 1. The focal technology cluster is
Smart Places that aims to address the challenges and opportunities of public institutions, utilities,
information and communication technology, logistics and transportation, financial and business
services, education, and the health sector. The leading team that performed the research and
knowledge sharing on SSEG integration is from the Energy Systems Research Group situated in the
CSIR Energy Centre. The stakeholder collaborating with the CSIR and involved in partially funding
the implementation of SSEG municipal business support is the South African−German Energy
Programme implemented by the German Development Corporation, GIZ.

Report for
Doc No: Rev Page 11 of 88
Figure 1: Technology sector clusters of the CSIR

CSIR Energy Centre value chain

Figure 2: Organisational structure of focal cluster, centre, and research group

DISCLAIMER
The information contained in this guide is for information purposes only. Users are advised to refer
to standards and guidelines mentioned in Section 7 for more detailed information. It is meant to serve
as a guide for users and is by no means prescriptive, mandatory, or legally binding.

Report for
Doc No: Rev Page 12 of 88
About the team
Mr Mpeli Rampokanyo is a Professional Electrical Engineer
from South Africa. He has a BSc degree in Electrical
Engineering from University of Cape Town and Master of
Engineering in Electrical Engineering from University of the
Witwatersrand. At the time of the compilation of this
document, he was appointed as a Principal Engineer in the
CSIR Energy Centre. He provides research leadership in
power system operations in terms of model development,
standards or codes, regulation and policy issues for the
current grid and a future energy-system with a high share of
Mpeli Rampokanyo renewables in support of the country’s energy goals. He is
a convenor of the CIGRE joint working group C2/C4.41:
Impact of high penetration of inverter-based generation on
system inertia of networks.

Fiona Oloo is a Candidate Research Engineer in the CSIR


Smart Places cluster working in the Energy Centre. Her
main responsibilities include energy modelling, power
systems analysis and energy system operation analysis.
She holds a Bachelor's in Engineering Science in
Biomedical Engineering and a BSc in Electrical Engineering
(hons), both obtained from the University of the
Witwatersrand. She started her career as a Graduate in
Training with the Aeronautical Systems research group in
CSIR Defence and Security, where her main responsibilities
involved gaining training towards competency to register
with the Engineering Council of South Africa. She then
moved to the Next Generation Enterprises and Institutions,
Fiona Oloo Advanced Internet of Things research group.

Report for
Doc No: Rev Page 13 of 88
Basetsana has a BSc degree in Electrical Engineering from
the University of the Witwatersrand. At the time of the
compilation of this document, she was seconded to the
Energy Systems research group of the CSIR’s Smart
Places Cluster, until June 2021. She is a Candidate
Electrical Engineer at the Department of Public Works and
Infrastructure. Her main responsibilities are in power system
analysis and developing energy efficiency solutions for
government departments. She is working towards
registering as a professional engineer with the assistance
of the CSIR.

Basetsana Molefyane

Lindokuhle is a Candidate Research Engineer at the CSIR


Energy Centre in the Smart Places Cluster. His main
responsibilities include energy and power systems
modelling, analysis and research with the purpose of aiding
and contributing to the field of power systems and energy
industry. He has a BSc degree in Electrical Engineering and
a Bachelor of Engineering Science in Biomedical
Engineering. His experience includes avionics design for
unmanned aerial vehicles, communications networks for
smart grids, and data extraction, processing and
visualisation of energy modelling data. He has gained skills
in power systems modelling, simulation and analysis and
Lindokuhle Mbatha project management.

Report for
Doc No: Rev Page 14 of 88
Stakeholders and roles
Stakeholders Role

Main implementing agency of South


African−German Energy (SAGEN) Programme
together with the GIZ that provides assistance to
South African municipalities.

GIZ are the leaders of SAGEN and collaborate


with several other stakeholders implementing the
programme.

Department of Mineral Resources and Energy of


South Africa is a collaborator with GIZ in the
implementation of the SAGEN Programme.

SALGA is an association of local South African


governments and a collaborator with GIZ in the
implementation of the SAGEN Programme.

SEA is a non-profit organisation promoting


equitable, low carbon, clean energy development
in urban South Africa and Africa. It played a review
role in this document.

Report for
Doc No: Rev Page 15 of 88
EXECUTIVE SUMMARY
South Africa’s power sector has seen a challenging number of years that threaten energy security
and economic prosperity of the country. Certain parts of the country do not have access to electricity
while regions that do have access have to deal with unreliability of supply and increasing electricity
costs. South Africa’s electricity is also predominantly fossil fuel based, which hinders South Africa’s
ambitions to meet the Paris Agreement obligations of reducing greenhouse gas emissions and
mitigate climate change. Given this situation, electricity consumers are turning to renewable energy
sources and particularly to distributed energy resources using Small-Scale Embedded Generation
(SSEG) technologies as this can provide a more affordable alternative to solely relying on the power
provided by their utility provider. These technologies allow consumers to manage the risk associated
with South Africa’s energy security. The South African Government is also encouraging the use of
renewable energy technologies to reshape the country’s future energy supply.
There is supporting data that demonstrate a significant increase in the installation of SSEG across
the country from 2016 to 2020, and this trend is expected to continue in the years to come. This can
be attributed to increases in grid-based electricity costs, supply unreliability and indirect grid
emissions while distributed energy technologies are typically renewable, are steadily becoming more
cost competitive and can supplement grid-based electricity or allow customers to go completely off-
grid. Of all the licensed municipal electricity distributors in South Africa, 34% allow for the installation
of SSEG, with 27% having official processes in place to do so. However, only 8.5% have an SSEG
tariff structure in place, but this is growing. Challenges identified in facilitating the process of
integrating SSEG into utility networks include the lack of technical skills to facilitate approval
processes, and a lack of understanding of the technical implications and revenue impacts of SSEG
on the municipal electricity distribution business.
The CSIR in collaboration with the Department of Mineral Resources and Energy, SALGA and GIZ
has engaged with metropolitan municipalities on understanding the impacts of SSEG on municipal
networks. The collaboration supported municipalities with the assessment of the technical impact of
SSEG integration into their networks. Seven municipalities showed interest in these topics, of which
three were evaluated for their readiness to facilitate SSEG interconnection. The evaluation focused
on existing business processes, availability of resources in terms of personnel, skills, and tools
(software), and the availability of relevant and accurate data. The outcome of the evaluation identified
gaps in terms of a lack of skills and capacity, and a lack of appropriate tools to conduct detailed
network assessments. Additional outcomes from the evaluation were issues identified by
municipalities concerning specifications and recommendations of guidelines that cover the
integration of SSEG into utility networks.
To address the gaps and concerns identified from the assessments, a set of specifications were
studied and reviewed. These included the NRS 097 series which outline specifications meant for use
in the electricity supply industry (ESI) and cover the topic of grid interconnection of embedded
generators. It includes a previously developed guideline by the MPE in 2018 which serves as a tool
for simple decision-making in SSEG integration.
Municipality engagements were conducted in the form of training and capacity building to provide
the engaged municipalities with the necessary information for carrying out simplified connection
assessments and detailed network analyses when required. Support tools developed by the CSIR
include an Excel-based simplified load flow assessment tool and a hosting capacity analysis tool
implemented on DigSILENT PowerFactory to assist municipalities with detailed assessments of
integrating SSEG into their networks.

Report for
Doc No: Rev Page 16 of 88
Investigations into the existing 350 kVA limit in the NRS 097-2-3 and standards for uninterrupted
power supply (UPS) connection were also conducted to address concerns raised by municipalities.
The first investigation included determining the basis of the 350 kVA limit and assessing if this limit
could be adjusted to accommodate increased SSEG capacity without the need to undertake detailed
studies. The second involved reviewing existing standards and specifications regarding UPS
connection and classification. Recommendations for amendments to the NRS specifications have
been proposed based on the investigations conducted to address municipalities’ concerns and
potentially update these in future based on the findings.

Report for
Doc No: Rev Page 17 of 88
1 Introduction

1.1 SSEG integration status in South Africa


1.1.1 South African electricity mix
With the South African energy supply predominantly being provided by fossil fuels, there is ample
room to expand the integration of renewable energy. This is shown in Figure 3 where the
predominant fossil fuel is coal that makes up ~82% of the electricity generation [6].

Figure 3: South Africa's electricity mix 2020 [6]

In line with the national commitment to ensure the transition to a low carbon economy, 17.8 GW of
the target set out in the Integrated Resource Plan (IRP) 2019, for 2030, is expected to arise from
renewable energy sources [7]. By 2019, the expectation was that 5 GW would be in operation by
2019 and an additional 2 GW by 2020 [7]. By the end of 2019, 4 054 MW of renewable energy
capacity was operational and by the end of 2020, 5 027 MW was in operation [6].

1.1.2 SSEG uptake statistics and challenges in South Africa


With approximately 3 million households in South Africa lacking access to grid-based electricity, non-
grid solutions have proven effective in the provision of lighting and small power needs but has shown
to be inappropriate in thermal energy provision for activities such as heating and cooking. The cost
of providing grid connection to many of these homes is increasing and more so if the areas served

Report for
Doc No: Rev Page 18 of 88
are more remote. It then becomes important to quantify off-grid and mini-grid opportunities to
accelerate energy access to more homes.
Distributed generation in the form of Small-Scale Embedded Generation (SSEG) that makes use of
various renewable energy sources, such as solar, wind and biomass as fuel, holds potential for
contributing to meeting the electricity needs in South Africa. SSEG refers to generation installed
behind the meter, synchronised with the distribution grid, up to a maximum capacity of 1 MW and is
installed on various properties such as residential, commercial, industrial or agricultural by electricity
customers [8]. In South Africa, 99% of the total 3 280 MW installed embedded generation is SSEG
that is <1 MW, with the remaining 1% embedded generation in the range 1−10 MW [8]. The
installation of SSEG is increasing considerably due to load-shedding concerns, increasing electricity
tariffs from South Africa’s main power producer (Eskom), while the technology costs of SSEG
decreases. Most SSEG installations have low-capacity factors and because of this and the fact that
these also tend to be intermittent in nature, must be controlled and integrated by smart technologies
[9]. Given the energy crisis experienced in South Africa, SSEG appears to be a viable option to
consider as it is a generation option that can be rapidly deployed.
The installation allocation as per the IRP 2019 allocates 500 MW per annum commencing in 2023
for embedded generation[10]. For 2019−2022, no specific capacity allocation is prescribed; instead,
the IRP 2019 provides for ‘an allocation to the extent of the short-term capacity and energy gap’
which is estimated at 2 000 MW[10].

60
56
50
Number of municipalities

44
40 41

34
30 31
29
25 25
20 21
17 18
12
10 10
5
3
0
Feb 16

Dec 16
Feb 17

Dec 17
Feb 18

Dec 18
Feb 19

Dec 19
Feb 20
Apr 16

Aug 16

Apr 17

Aug 17

Apr 18

Aug 18

Apr 19

Aug 19

Apr 20
Oct 16

Oct 17

Oct 18

Oct 19
Jun 16

Jun 17

Jun 18

Jun 19

Jun 20

Allow SSEG Official application process in place SSEG tarriff implemented

Figure 4: National uptake of SSEG in municipalities South Africa [8]

Figure 4 illustrates the uptake of SSEG from municipal distributors in South Africa into their networks
showing steady increase in allowing SSEG to be connected to the distribution system, accompanied
by official application processes and SSEG tariffs implemented. From February 2016 to June 2020,

Report for
Doc No: Rev Page 19 of 88
the municipalities that allowed SSEG uptake increased by just over five times, the official processes
in place by 15 times, and SSEG tariffs implemented by six times. This has been influenced by efforts
to build skills capacity, providing support to municipalities for integrating SSEG, legislative
frameworks that have facilitated SSEG uptake, and municipalities showing increasing willingness in
partaking in the energy transition in South Africa.
Figure 5 shows an overview of SSEG uptake by province in South Africa as of 2020. Things to note
include:
• The Western Cape allows the most SSEG installations, has the highest number of
municipalities with official application processes including the implementation of SSEG tariffs.
• KwaZulu-Natal has the same number of municipal electricity distributors as the Western
Cape, but it is noted that the number of municipalities allowing SSEG installations is 18 less
than the Western Cape.
• Of the number of municipal electricity distributors in the Western Cape, 88% allow SSEG
installations and this is predominantly because of the provincial government providing
support for SSEG.
• For the Free State, the number of municipalities allowing SSEG installations, municipalities
that have official application processes and those with SSEG tariffs, is shown as zero in
Figure 5, but it must be noted that this is not necessarily the case, no data could be recorded
for these categories.
• An observation for each of the provinces, excluding the Free State in this case due to the
unavailability of data, notes that the number of municipalities with SSEG tariffs is fewer than
the actual number that allows SSEG installations. This can be attributed to the fact that there
are uncertainties regarding the National Energy Regulator of South Africa (NERSA)
application/approval processes and billing system integration challenges as well as internal
municipal revenue concerns. Additionally, NERSA requires Cost of Supply studies to
accompany new SSEG tariff applications and some municipalities are yet to complete this.
South Africa has 189 licensed [11]. Overall, only 34% of municipalities are licensed distributors
allowing SSEG, 27% are licensed with official application processes, and 19% are licensed with an
SSEG tariff structure in place.

Report for
Doc No: Rev Page 20 of 88
Number of
180 municipal
165
170 electricity
160 distributors in
150 province
140
130 Number of
120 municipalities
110 allowing SSEG
Number

100 installations
90
80
70 56 Number of
60 municipalities with
50 44
official application
40 31 processes
22 25 24 2522
30 17 16 1819
20 14 13
662 9 65 9
10 000 432 321 1 443 43 220 Number of
0 municipalities with
SSEG tariffs

Province

Figure 5: Provincial SSEG uptake summary [8]

Figure 6 shows the estimated registered SSEG capacity[8]. Gauteng, despite having only four
municipalities that allow SSEG integration, has the highest proportion of registered capacity in South
Africa (majority of the municipalities are in more densely populated metropolitan areas) at 46% of a
total of registered 282.2 MW countrywide.

Report for
Doc No: Rev Page 21 of 88
Eastern Cape
4% Free State
0%
Western Cape
23%

North West
6%

Northern Cape Gauteng


1% 46%

Mpumalanga
6%

Limpopo
1%
KwaZulu-Natal
13%

Eastern Cape Free State Gauteng KwaZulu-Natal Limpopo


Mpumalanga Northern Cape North West Western Cape

Figure 6: Estimated capacity (MW) of registered SSEG system in South Africa [8]

Municipalities noted various areas where support is required with grid impacts studies being the
predominant support required. Other support areas include (from most required to least
required):
• Tariffs for SSEG
• Management and council briefings
• Technical support and training (general)
• SSEG application processing
• Bi-directional metering/billing
• Commissioning and inspections of SSEG
• Policy and documentation development
• Approaches to illegal SSEG installations
• Exchanges between municipalities
• SSEG database and registration
• Customer awareness.

Report for
Doc No: Rev Page 22 of 88
Political
11%

Distribution
network capacity Technical capacity
13% 46%

Approval
processes
30%

Technical capacity Approval processes


Distribution network capacity Political

Figure 7: Main challenges in SSEG integration into municipal distribution [8]

A brief description of the challenges in include[8]:


• Technical capacity: This revolves around shortage of staff and staff being unfamiliar with the
SSEG assessment and technologies. SSEG support programmes should assist in alleviating
this challenge.
• Distribution network capacity: This challenge refers to understanding/knowledge of the
distribution network and the available capacity to integrate SSEG, i.e. the SSEG integration
that is possible is limited by the hosting capacity of the distribution network.
• Approval processes: This challenge refers to a lack of formal SSEG approval processes.
Without formal processes, the integration of SSEG cannot be done formally and thus
registration and/or licensing with NERSA may not occur.
• Political: Political challenges faced in accepting SSEG into municipal networks may be due
to misunderstandings regarding technical and revenue impacts of SSEG on distribution
networks.
Figure 7 highlights the main challenges faced in accepting SSEG connection into municipal
distribution networks as per a municipal survey mentioned by SALGA [8]. Important points to note
are:
• Lack of technical capacity, which is the main challenge for 46% of municipalities, followed by
30% whose main challenge is a lack of formal approval processes. These are the two
predominant challenges that must be addressed first and foremost. Concerted efforts are
needed to capacitate municipalities with technical staff and further establish formal approval
processes to improve the streamlining of SSEG uptake.

Report for
Doc No: Rev Page 23 of 88
• The challenge around available capacity on the distribution grid to accommodate SSEG
(13%) is nearly the same percentage as the political challenges (11%). Political challenges
may point to misunderstandings around the impacts (technical and revenue) of SSEG on
municipal distribution networks, while the challenge around distribution network capacity
refers to the limitation of the current distribution network to safely host increasing SSEG.
Network improvement/s could assist in alleviating this challenge.

1.1.3 Consideration of the impacts of COVID-19 relating to SSEG uptake


Just as it has in many other sectors in the globe, the COVID-19 pandemic has caused many
challenges in the energy sector, particularly around energy usage and CO2 emissions[12],[13]. With
mandatory lockdowns implemented across the globe during the first quarter of 2020, which reduced
economic activity and mobility, energy demand reduced by 3.8 % around the world compared to the
first quarter of 2019[13]. This is the first decline in demand noted in the past 70 years[13]. Some
impacts and their possible effects, including the drop in electricity demand and the inability of some
customers to pay for electricity because of job losses in the pandemic, have placed financial burden
on electricity providers [14]. Thus, there is a possible need to analyse the effects of SSEG uptake
more carefully and thus SSEG tariff structures on the revenue of municipalities in the short term.
Renewables have shown to be the more resilient energy technology compared to fossil fuels during
the pandemic and since resilience is a key requirement for power systems, this may likely further
encourage SSEG uptake in distribution networks[14]. The disruption on power sector supply chains
due to COVID-19 [14] is one aspect that hinders the sourcing of equipment for SSEG uptake, thus
potentially slowing down SSEG deployment. Currently, short-term impacts are increasingly being
monitored and are becoming clearer, but longer-term impacts are yet to be determined as the world
moves to a post-pandemic state.

1.1.4 Existing practices and policies around SSEG integration


Licensing and registration for SSEG connection is required to ensure safe connection, and that
compliance with safety regulations is accounted for (for staff and customers). It is also required to
ensure integrity of the grid infrastructure and to allow for proper planning of future upgrades and
maintenance of the grid. Existing barriers that may hinder/slow down deploying SSEG include
uncertainty around stakeholders and the project risk associated with a regulatory landscape that is
indistinct [15]. If installed, SSEG systems are discovered to be incorrectly licensed, and registered
customers can be liable to pay a service fee for the disconnection of unauthorised SSEG
connections. Reconnection will only occur if the relevant authority authorises the connection when
the correct procedure is followed for SSEG connection.
Currently, under South African policy and legal framework, operation of the distribution, generation
or transmission facility must be licensed or registered with NERSA. All generation that is grid-tied
must be registered with the relevant transmission/distribution grid operator. The three possible
processes, depending on the system size, are:
1. Registration
2. Licensing
3. Ministerial determination.
The licensing and registration for various current SSEG system sizes are shown in Table 1. SSEG
in the range 100 kW−1 MW must be registered with NERSA, but licensing is not a requirement, while

Report for
Doc No: Rev Page 24 of 88
in the range 1 MW−10 MW, licensing from NERSA is required. SSEG under 10 MW does not need
to be part of a ministerial determination, while higher than 10 MW must be licensed and be part of a
ministerial determination and national procurement. By law, registering and licensing lies with
NERSA and not municipalities. Following an announcement by the Presidency in South Africa in
June 2021, an amendment of Schedule 2 of the Electricity Regulation Act has been issued in August
2021 in which energy generation projects up to 100 MW are exempt from requiring NERSA licensing,
but registration and grid compliance would still be necessary [16][17].

Table 1: Licensing and registration details for various sizes of embedded generation
systems

Embedded NERSA Eskom/Municipality


generation size
Registration Licensing Application for
required? required? connection


< 100 kW
✖1 ✖

100 kW–1 MW ✔ ✖ ✔

1 MW–100 MW ✔ ✖ ✔

> 100 MW ✔ ✔ ✔
1There must be an existing point of connection and a register of these installations must be kept by the local distribution
utility. The local distribution utility must prescribe to the conditions of connection.

Figure 8 shows the registration and approval process for a generation connection [18]. The process
begins with applying at the local distributor (municipality or Eskom) for connection to the grid. After
this, an application is made to NERSA to register the generation plant which requires a letter from
the local electricity distributor that confirms that connection permission is granted.
One aspect that is a challenge for safely increasing SSEG integration on the distribution grid is the
lack of clarity/data on the existing unregistered SSEG. There is a considerable number of SSEG
installations not registered in South Africa. A most recent estimate is currently being investigated by
ongoing studies. For various reasons, including the enabling of better planning of SSEG integration
onto the national grid, these systems must be registered.

Report for
Doc No: Rev Page 25 of 88
Phase 1: Application to local Phase 2: NERSA
distributor for connection registration

Figure 8: Registration and approval process for SSEG

Report for
Doc No: Rev Page 26 of 88
The standard conditions (technical standards and/or specifications) that facilities registered with
NERSA must comply with are as shown in Figure 9 [19].

Grid interconnection of
Embedded Generation SANS 10142 (Parts 1 to
NRS 097-1: 4): The wiring of premises;
NRS 048: Electricity
Provides practical contains the most
Supply−Quality of Supply
guidelines to interface applicable requirements of
in the electricity supply
SSEG to utility network. electrical installations
industry.
NRS 097-2: SSEG creates additional
Requirements for requirements.
embedded generation.

South African Grid Code


Requirements for NRS 057/SANS 474:
Renewable Power Plants Code of Practice for
NRS 047: Electricity Electrical Metering.
Specifies the minimum Supply−Quality of Service Prepared for application in
technical requirements for in the electricity supply the South African
grid connection of RPPs to industry. electricity supply industry
South African electricity to establish and promote
transmission/distribution uniform requirements.
systems.

Figure 9: Standard conditions with which NERSA-registered facilities must comply

Report for
Doc No: Rev Page 27 of 88
Report for
Doc No: Rev Page 28 of 88
Identified gaps in municipal readiness – overview
From a total of seven municipalities that showed interest in receiving support with SSEG grid impact
studies, three municipalities were selected for support. A gap analysis was performed using four
metrics to measure the readiness of municipalities for SSEG integration. These metrics were:
• Business processes − The focus is on the availability of SSEG application forms, the policies
and application procedures, and responsibility of municipal personnel.
• Skills and capacity of personnel − The focus is on the technical skills and the capacity of
personnel to conduct technical grid impact assessments when SSEG is to be integrated.
• The availability of data − The focus is on the data available when a technical assessment
of SSEG connection is required.
• The availability and application of tools appropriate for assessment of SSEG
integration – The focus is on resource availability and adequacy of tools.
Upon completion of the gap analysis with the three municipalities, the following gaps were
identified:
• Shortage of personnel skills and capacity, i.e. a lack of dedicated staff to perform grid
impact studies.
• Limitations of the NRS 097 in terms of the guidance provided only for SSEG connections
≤350 kVA.
• Absence of appropriate tools to perform detailed grid impact studies.
• Contradiction between existing policy specifications and technical requirements around
grid-tied system definitions.
Municipalities also outlined their goals in terms of the metrics used in the gap analysis:
• Business processes: For their business processes, they desire to have processes clearly
documented (application to commissioning of SSEG), including grid impact study
processes and standardised information exchange within municipal departments.
• Personnel and skills capacity: Personnel with adequate time, resources and skills to
assess SSEG applications.
• Availability of data: To have accurate spatial data, accurate technical specifications, load
profiles for medium voltage (MV) and low voltage (LV) networks, and to have accurate data
about existing SSEG connections.
• Resource availability and adequacy: Have access to screening criteria and the ability to
carry out hand calculations for SSEG applications as well as power system simulation
software for detailed SSEG grid impact studies.

Section 2 covers the identified gaps in municipal readiness in detail.


(Next section: Guidelines and capacity building – overview)

Report for
Doc No: Rev Page 29 of 88
2 Identified gaps in municipal readiness for SSEG integration
Via the third South African−German Energy Programme, the CSIR as an implementing partner
together with GIZ, in collaboration with the Department of Mineral Resource and Energy and the
South African Local Government Association (SALGA), invited metropolitan municipalities to apply
for support to assess technical impacts of SSEG on municipal distribution networks. A total of seven
municipalities showed interest in receiving support with SSEG grid impact studies and three
municipalities were selected.

The municipalities selected for support were:


• Nelson Mandela Bay Municipality
• City of Cape Town
• eThekwini Municipality.

Four metrics were considered to measure the readiness of the municipalities to integrate SSEG in
their distribution networks. The first metric was business processes in which the focus was on the
availability of SSEG application forms, the policies and application procedures and responsibility of
municipal personnel. The second metric was the skills and capacity of personnel. This metric focused
on the technical skills and the capacity of personnel to conduct technical grid impact assessments
when SSEG is to be integrated. The third metric is the availability of data. This metric focused on the
data available when a technical assessment of SSEG connection is required. The fourth metric
involved the availability and application of tools appropriate for assessment of SSEG integration.

Gaps were identified to develop mitigating solutions to improve the readiness of the municipalities.
These gaps are detailed in sections 2.1, 2.2, 2.3 and 2.4. Section 2.1 details a gap in terms of
shortage of personnel skills and capacity to undertake SSEG grid impact studies. Section 2.2
highlights the absence of tools to perform detailed grid impact studies. Section 2.3 focuses on the
limitation of the NRS 097-2-3 simplified connection criteria (a South African guideline for simple
assessment of a possible SSEG connection to a network). Section 2.4 highlights contradictions
between existing municipal specifications versus technical requirements. Section 2.5 details the
desired goals of municipalities in terms of the metrics.

2.1 Shortage of personnel skills and capacity


One of the gaps highlighted was the shortage of dedicated personnel to perform simple SSEG grid
impact studies. For some municipalities, existing staff performed the work in addition to other job
requirements. In the worst case, the available personnel did not have the requisite skills, an
understanding of simplified connection criteria, performing simple hand calculations or detailed grid
impact studies using power system analysis tools. To work around not having dedicated personnel,
external service providers would be appointed to perform the technical assessments. However, even
with this approach, no formal agreement existed between the municipality and the service provider.
The formal agreement would be between the SSEG applicant (customer) and the service provider.

Report for
Doc No: Rev Page 30 of 88
2.2 The absence of appropriate tools to perform detailed grid impact
studies
There were instances where municipalities did not have appropriate tools to perform detailed grid
impact studies, such as power system analysis tools. The municipality would either not have
licence/s for the typically utilised commercial tools, or the licence would exist but not have appropriate
functionality. Thus, for detailed studies, the municipality would require the assistance of an external
service provider to perform the detailed studies.

2.3 Limitations of the NRS 097-2-3 Simplified Connection Criteria


Municipalities with dedicated personnel who understand the South African specification for simplified
connection criteria of SSEG (NRS 097-2-3) [20] indicated that guidance and assistance on
enhancing the method was needed, i.e. to not only cover generators ≤350 kVA. The NRS 097-2-3 is
for low voltage (LV) connected generators and establishes whether SSEG can be connected without
performing detailed studies. Maximum permissible generation size depends on the type of LV
network (either shared feeders or dedicated feeders) and the notified maximum demand (NMD). If
connection criteria are not met, then detailed network studies must be performed. Unlike detailed
studies, the methods in NRS 097-2-3 are relatively quicker to perform. However, for accuracy,
detailed studies are a better option, but this is often not a feasible option for many municipalities
given their severe capacity constraints, hence the NRS 097-2-3 is essential for them. Most
municipalities would opt for the simplified connection criteria because it allows for faster processing
of applications.

2.4 Existing specifications and technical requirements


For some South African metropolitan areas, there is a fast pace increase in customers installing
SSEG. As a result, municipalities need to ensure that the existing systems and processes
accommodate for these changes. Municipalities not experiencing the capacity gaps, highlighted
more technical issues that may arise when there is an influx in the activity of customers installing
SSEG. One such problem concerns a possible contradiction between existing specifications and
technical requirements.
The problem was that customers who elected to install a particular configuration of SSEG declared
to be off-grid where, in fact, they were grid-tied and would consume power from the network. The
specific municipality allows for a standby passive uninterrupted power supply (UPS) system to be
installed that meets a particular criterion. This type of system is then categorised as an off-grid
system, however, the normal operation is grid-tied with load being supplied by the grid as the primary
power source, while outside normal operations the load is supplied from the UPS system operating
in a new mode (stored-energy mode). This type of system can be declared off-grid if it is equipped
with a suitably inter-locked change-over switch. From a technical perspective, a standby UPS system
does not meet the requirements of an off-grid system, defined as “the system is physically
disconnected and electrically isolated and can never be connected to utility grid". The concern arises
when customers who are off-grid are not limited to designing systems that comply with the
requirements of grid-tied systems. Such cases would be designing a system that is greater than the
After Diversity Maximum Demand (ADMD) and equal to the NMD, for which the network is not
designed. Should there be an increased number of customers installing these types of systems,
there will be adverse effects on the network when the loads of these customers consume more than
the designed ADMD simultaneously. For this reason, the contradiction between the policy agreement

Report for
Doc No: Rev Page 31 of 88
and technical requirements will have adverse effects on the network when more customers install
these systems.

2.5 Desired goals for municipal SSEG integration


The desired goals for municipalities for each of the above-mentioned metrics are detailed in Table
2. It outlines what goals the municipalities would like to achieve for each of the metrics.

Table 2: Future goals in terms of the metrics used to determine municipal readiness for SSEG
integration

Desired Future Goals (For SSEG grid impact assessment)


Business ➢ Clearly documented process from application to commissioning.
processes ➢ Application that captures all relevant information.
➢ Clearly documented grid impact studies process.
➢ Standard manner of information exchange across internal municipal
departments.
Personnel and ➢ Human resources with adequate time to assess SSEG applications, these
skills capacity could be resources who are dedicated to perform grid impact assessment on
SSEG applications rather than doing the work as an add-on to their official
job descriptions.
➢ Resources with adequate skills to assess SSEG applications. They should
have the ability to perform hand calculations, understand the simplified
connection criteria or detailed studies. Added to that, they should be able to
comprehend when each method of assessment is applicable.
Availability ➢ Accurate spatial data for MV and LV networks.
of data ➢ Accurate technical specifications/ parameters of MV and LV networks.
➢ Accurate load profiles for all customer classes at MV and LV.
➢ Accurate data regarding SSEG that is already connected to the network.
Resource ➢ Screening criteria for assessing SSEG applications.
availability and ➢ Hand calculations for assessing simple SSEG applications.
➢ Power system simulation software to conduct detailed SSEG grid impact
adequacy studies.

Report for
Doc No: Rev Page 32 of 88
Guidelines and capacity building – overview

Several guidelines are currently available to assist in assessing SSEG suitability for integration into
the power grid. These include the NRS 097 documentation that contains specifications for the
electricity supply industry (ESI) and the Moeller & Poeller Engineering (MPE) guideline, which is a
recommended practice that can be used to assess connection of small generators to LV/MV grids
by using hand calculations.
NRS 097 series
The series has two parts to it:
• NRS 097-1: Distribution standard for the interconnection of embedded generation – Provides
connection specification for embedded generators rated 100 kVA or higher for connection to
MV and high voltage networks.
• NRS 097-2: Small-scale embedded generation – Provides specification for SSEGs rated
1 000 kVA or less connected to the LV network.
Focusing on the LV network, the NRS 097-2 is separated into four sections as follows:
• NRS 097-2-1: Utility interface – Outlines SSEG integration requirements at point of
connection.
• NRS 097-2-2: Embedded generator requirements – A guideline for SSEG compliance tests.
• NRS 097-2-3: Simple utility connection criteria for LV connection generators – Provides
guidelines on design aspects for SSEG interconnection into the LV network.
• NRS 097-2-4: Procedures for implementation and application – Provides information on
connection agreements, application forms, and data requirements.
Of the four sections, two are currently available to the public – the NRS 097-2-1 and NRS 097-2-3 –
details of which are provided in section 3.1.2 and section 3.1.3, respectively.

Report for
Doc No: Rev Page 33 of 88
The MPE guideline
• Guideline covers SSEG integration with capacities: <100 kW and between 100 kW and 5MVA.
• Covers simplified methods and points to the conditions where detailed analysis is required to
provide further guidance on a potential SSEG integration.
• Provides guidance to determine the impact of connecting generation to the MV grid.

Section 3 covers guidelines and capacity building in detail.


(Next section: CSIR Simplified load flow tool – overview)

Report for
Doc No: Rev Page 34 of 88
3 Guidelines and capacity building
Considering the gaps identified in South African municipalities regarding the ability to facilitate the
integration of SSEG into low-voltage (LV) networks, the SSEG Municipal Grid Impact Study work
package, under the South African−German Energy Programme, aims to capacitate municipalities
with the necessary information and tools for the job. This involved studying and reviewing guidelines
that speak to SSEG interconnection to utility grids and engaging with municipalities for capacity
building and information dissemination in this regard.
The guidelines reviewed include the NRS 097-2-1 and NRS 097-2-3, which outline specifications
that can be utilised as requirements and recommendations to follow when assessing SSEG for their
suitability to interface with the utility network. The Moeller & Poeller Engineering (MPE) guideline
[21] developed in 2018 was also considered as it is an alternative to the NRS 097-2-3 for a simplified
decision-making process and provides information for performing detailed studies when required
based on the evaluation process given in the NRS 097-2-3.
The CSIR’s implementation support provided to municipalities involved providing guidance to
municipal personnel on simplified assessment for SSEG integration, and when and how to perform
detailed network analyses if required. To aid municipal personnel in performing detailed network
studies, an Excel-based simplified load flow assessment tool and a hosting capacity tool hosted on
DigSILENT PowerFactory were developed.
The following sub-sections provide information on the NRS 097 series and the MPE guideline, which
speak to the specifications, requirements, and recommendations for integrating SSEG into LV
networks. The section then follows with the engagements conducted with municipalities on
guidelines and processes for SSEG integration. The tools meant to help facilitate detailed analysis
are then described in terms of their capabilities and functionalities. An analysis and investigation to
the issues regarding the NRS 097-2-3 350 kVA limit and UPS connection raised by municipalities
follows, with recommendations made to address these concerns.

3.1 NRS 097


The NRS 097 is a set of specifications meant to serve the electricity supply industry (ESI) and covers
the topic of grid interconnections of embedded generators. There are two parts to the NRS 097 that
speak to two separate size categories of embedded generation. These are:
• Part 1: Distribution standard for the interconnection of embedded generation, is still under
development and is intended to outline the specification for the minimum technical and
statutory requirements for the connection of embedded generators to the medium voltage
(MV) and high voltage utility distribution network [22]. The specification would apply to
embedded generators larger than or equal to 100 kVA. A significant part of the requirements,
however, can be found in the grid connection code for renewable power plants [23].
• Part 2: Small-scale embedded generation, sets out the technical requirements for the utility
interface, the embedded generator and/or system, and the utility distribution network with
respect to embedded generation [24]. The specification applies to SSEGs smaller than or
equal to 1000 kVA connected to the LV network.
This document is intended to provide information surrounding the integration of SSEGs into LV
networks. As such, the main focus in the discussion to follow will be on part 2 of the NRS 097 series
(i.e., NRS 097-2). Section 3.1.1 provides a background into the NRS 097-2, with brief descriptions

Report for
Doc No: Rev Page 35 of 88
of sections 1 to 4 of the NRS 097-2. Of the series, information of the NRS 097-2-1 and NRS 097-2-
3 will be outlined.

3.1.1 NRS 097-2


The main objective of the NRS 097-2 is to address the lack of pre-approved, generic standards for
utility engineers and system promoters to apply in designing and approving an SSEG interface for
the utility. The NRS 097-2 is divided into four sections which are described as follows [22]:
• Section 1: Utility interface – This document outlines the requirements at the point of
connection (POC) and prioritises the safe interconnection of SSEG. The specification is
similar to various international standards such as the IEC 61727, IEEE 1547, EN 50438,
VDE-AR-4105, G83 and G59, with very minimal changes made to suit South African
networks.
• Section 2: Embedded generator requirements (to be developed in the future) – This
document is intended to provide guidance on tests required to prove compliance to section
1.
• Section 3: Simple utility connection criteria for LV connection generators – This document
provides guidelines for utilities on planning of design aspects of LV networks to facilitate
application for SSEG interconnection.
• Section 4: Procedures for implementation and application (to be developed in the future) –
This document should provide details of process-related aspects surrounding SSEG
interconnection and will include information on connection agreements, application forms,
and data requirements.
Two of these four sections are currently available, with the latest edition of Section 1 (NRS 097-2-1)
published in 2017 and Section 3 (NRS 097-2-3) published in 2014.
These two sections of the NRS 097-2 are discussed in more detail in 3.1.2 and 3.1.3.

Report for
Doc No: Rev Page 36 of 88
3.1.2 The NRS 097-2-1: Utility interface

The NRS 097-2-1 was initially published in December 2010 in response to the many requests and
enquiries received by distribution licensees regarding the connection of photovoltaic (PV) panels and
inverters in parallel to the utility network [22]. The latest and revised edition was later published in
2017 due to significant international developments made in the subject matter [22]. The NRS 097-2-
1 derives most of its specification largely from the IEC 61727 and the VDE-AR-N 4105, with additions
from various other international standards [22] [24]. The document aims to be technology agnostic,
covering energy sources associated with static power converter technologies – such as solar power,
wind power, micro-hydro power, pico-hydro power, battery storage energy and fuel cells – and other
conversion technologies, which include synchronous and induction generators [24].
This section of the NRS 097-2 covers embedded generation systems of nominal generation capacity
of less than or equal to 1 000 kVA, connected to single-phase, dual-phase, or three-phase LV
networks [24]. However, with the announcement made by the President of South Africa, Mr Cyril
Ramaphosa, on 10 June 2021 [25], which stated that an amendment will be made to Schedule 2 of
the Electricity Regulation Act to increase the NERSA licensing threshold for embedded generation
projects from 1 MW to 100 MW, the 1 000 kVA value may be revised in the next iteration of the NRS
publication.
The NRS 097-2-1 document has four sections. The first outlines the Scope of the document, which
in brief states that the document covers practical specification for utilities to facilitate the integration
of embedded generation into LV networks [24]. The scope does not, however, include requirements
for fuel sources and/or converters of alternate energy sources [24]. The second section provides a

Report for
Doc No: Rev Page 37 of 88
list of Normative References used to compile the specification outlined in section 1 of the NRS 097-
2. The list of Terms, definitions and abbreviations used in the document is given in section three.
Section four is where the bulk of the material sits as it covers the Requirements necessary for SSEG
interconnection to the utility. The Requirements section is divided into three parts, i.e. Utility
compatibility, Safety protection and control, and Metering. The content to follow is a brief discussion
of the parts that form the Requirements section.

3.1.2.1 Utility compatibility


The requirements under utility compatibility speak to the technical issues and responsibilities related
to integrating SSEG to the utility network. These requirements are related to SSEGs that fall under
Category A in accordance with the Renewable Power Plant Grid Code [23]. SSEGs under Category
A are further divided into subcategories based on generation capacity size as follows [24]:
• Category A1: 0 kVA – 13.8 kVA; rated power of embedded generators in this range is
inclusive of 13.8 kVA.
• Category A2: 13.8 kVA – 100 kVA; embedded generators with rated power greater than
13.8 kVA but less than 100 kVA.
• Category A3: 100 kVA – 1 MVA; rated power of embedded generator ranges from 100 kVA,
but less than 1 MVA.
Most requirements cover all SSEG systems, however, some are related to generation capacity sizes
of specific ranges. These are discussed in more detail in section 3.1.2.1.3. Further design
requirements related to capacity size of embedded generators are explored in NRS 097-2-3.

3.1.2.1.1 General Requirements under the NRS 097-2-1


All embedded generators are required to be wired permanently in accordance with the SANS 10142-
1 standards [24]. Embedded generators supply power to both the local loads and, potentially, to the
utility. As a result, the parameters of power quality delivered – which include voltage, flicker,
frequency, harmonics and power factor – should adhere to practices and standards outlined in the
NRS 048-2 [26]. The power quality should be determined and maintained by the utility, thus SSEG
should not attempt to regulate nor cause significant disturbances or deviations of the power quality
parameters from nominal or statutory conditions [24]. This requires that power quality parameters be
measured at the POC, and in the event that these parameters deviate from statutory conditions,
embedded generators may be required to disconnect from the utility [24]. Some of the critical
standards for the power quality parameters are summarised in

Report for
Doc No: Rev Page 38 of 88
Table 3 to Table 8.

Report for
Doc No: Rev Page 39 of 88
Table 3: Voltage requirements for grid-connected SSEG

Voltage requirements

Normal voltage operating ➢ 85% ≤ V ≤ 110%


range

Maximum direct current ➢ 1 000 V; this is the voltage on the DC side of an inverter, for
(DC) voltage instance when no load is taken and maximum source energy
is available (e.g. peak solar radiation for PV panels)

Maximum voltage change ➢ Generation rejection (i.e. tripping of SSEG while generating
at full capacity) should result in a voltage change of less than
3% at POC. This is in accordance with VDE-AR-N 4105.

Voltage unbalance ➢ Maximum unbalanced generation of 4.6 kVA for single- and
dual-phase embedded generation. Units larger than 4.6 kVA
should be split evenly between available phases.
➢ Voltage unbalanced contribution should be capped at 0.2%
for three-phase generators connected to a network with
impedance equal to that of the reference.

Report for
Doc No: Rev Page 40 of 88
Table 4: Frequency requirements for grid-connected SSEG

Frequency requirements

Normal frequency ➢ 47 Hz ≤ 𝑓 ≤ 50.5 Hz; in this range embedded generation


operating range systems shall operate normally.

Under-frequency ➢ When the utility frequency falls below 47 Hz, embedded


generator shall disconnect from the utility network within
0.2 s.

Over-frequency ➢ For frequencies between the range 50.5 Hz ≤ 𝑓 ≤ 52 Hz, the


embedded generation system shall control the output power
as a function of the active power available at a gradient of
50% per Hertz as shown in Figure 10 until the frequency has
stabilised below 50.5 Hz for at least 4 s.
➢ When utility frequency exceeds 52 Hz for more than 4 s, the
embedded generator should discontinue energising the utility
within 0.5 s.

Table 5: Harmonic distortion requirements for grid-connected SSEG

Harmonics and waveform distortion requirements

Harmonic distortion ➢ Total harmonic distortion should be limited to 5%.


➢ Harmonic and inter-harmonic current distortion should
conform to the relevant emission limits outlined in the
IRC 61727 and reproduced in Table 9.

Commutation notches ➢ Relative depth of commutation notches due to line-


connected inverters should be limited to 5% of the nominal
voltage at the POC for any operational mode.

Report for
Doc No: Rev Page 41 of 88
Table 6: Flicker requirements for grid-connected SSEG

Flicker requirements

Generation flicker ➢ When connected to a network impedance equal to the


reference impedance used during certification, SSEG
generation flicker levels are limited to the following:
o Short-term flicker severity (Pst) = 0.35
o Long-term flicker severity (Plt) = 0.30.
➢ The utility is expected to plan connections in line with
acceptable flicker limits.

Table 7: Power factor requirements for grid-connected SSEG

Power factor requirements

Power factor ➢ The power factor for static power converter embedded
generators, synchronous generators and asynchronous
generators categorised under sub-categories A1 and A2
should be above 0.98.
➢ For static converter embedded generators, synchronous
generators and asynchronous generators under sub-
category A3, the power factor should be above 0.95.
➢ Asynchronous generators without power factor control
capabilities should be able to reach required power factor
range within 60 s.

Report for
Doc No: Rev Page 42 of 88
Table 8: Information of reference source impedance, fault levels, and current injection requirements
for grid-connected SSEG

Reference source impedance, fault levels, and current injection requirements

Short-circuit level ➢ The minimum network strength for generator testing and
design under worst-case conditions are:
o Source impedance (Z_source) = 1.05 + j 0.32 Ω; i.e.
▪ I_SC = 210 A
▪ S_SC = 146 kVA (three-phase)
➢ Generator short circuit contribution is limited as follows:
o Synchronous generators: 8 times the rated current
o Asynchronous generators: 6 times the rated current
o Generators with inverters: 1 times the rated current

DC current injection ➢ DC current injected into the utility a.c. interface should not
exceed 0.5% of the rated a.c. output current over any 1-
minute period under any operational condition.

Table 9: Maximum harmonic current distortion as percentage of rated current (from NRS 097-2-1)

Harmonic Order (h) h<11 11≤h<17 17≤h<23 23≤h<35 35≤h

Percentage of rated current (odd harmonics) 4.0 2.0 1.5 0.6 0.3

Percentage of rated current (even harmonics) 1.0 0.5 0.38 0.15 0.08

Percentage of rated current (inter-harmonics) 0.1 0.25 0.19 0.08 0.03

Total Demand Distortion = 5%

NOTE 1: Even harmonics are limited to 25% of the odd harmonic limits.
NOTE 2: Inter-harmonic are limited to 25% of the odd harmonic limits and adjusted for the 200 Hz band
measurement required by IEC 61000-4-7, except for the lower frequencies where the flicker contribution
is more likely.
NOTE 3: Total Demand Distortion = Total Harmonic Distortion

Report for
Doc No: Rev Page 43 of 88
Power curtailment during over-frequency
110
% of Power output when f > 50.5 Hz

100
90
80
70
60
50
40
30
20
10
0
48 48.5 49 49.5 50 50.5 51 51.5 52 52.5 53 53.5
System frequency [Hz]

Figure 10: Power curtailment during over-frequency [24]

3.1.2.1.2 Electromagnetic compatibility


Electromagnetic compatibility (EMC) is a system’s or equipment’s ability to operate satisfactorily in
its electromagnetic environment without causing electromagnetic disturbances [24]. Standards for
EMC define the terms, rules, test methods, emissions limits and immunity levels for EMC [27]. The
following requirements regarding EMC are outlined in the NRS 097-2-1 [24]:
• The standard for unintentional emissions from generating equipment can be found in the
SANS 50065-1 and SANS 211 (otherwise referred to as the CISPR11), which cover the
frequency ranges 3 kHz to 148.5 kHz and above 148.5, respectively; emissions should
comply with the limits for Class B group 1 equipment outlined in SANS 211 (CISPR11).
• The conducted emissions requirement applies to all ports or connections to the utility supply,
and covers any purpose of connection, including monitoring, communication, or power
transfer.
• Under conditions of electromagnetic interference susceptibility, units shall be fail-safe, i.e.
deviations from intended performance should conform to all relevant specifications, both in
terms of the safety (i.e. disconnection) of and impact on the network.
• Despite this, should there be any interference to existing or new ripple control, building
management system equipment and/or other power line carrier-based communication, the

Report for
Doc No: Rev Page 44 of 88
customer responsible for the embedded generator should remedy the situation to prevent
further interference based on agreement with the utility or other affected party.

3.1.2.1.3 Requirements based on size


There are other requirements based on generation capacity size in addition to the voltage unbalance
requirements mentioned in

Report for
Doc No: Rev Page 45 of 88
Table 3. These requirements are as follows [24]:
• The maximum size of an embedded generator is limited by the rating of the supply point on
the premises.
• Units larger than 13.8 kVA shall be of balanced three-phase type unless only a single-phase
network supply is available, in which case NRS 097-2-3 recommendations can be applied
based on the notified maximum demand (NMD).
• A customer with a multiphase connection intending to connect an embedded generator larger
than 4.6 kVA should split the connection between available phases in a balanced manner.
• Additional requirements may apply for generation systems larger than 100 kVA, such as the
capability of receiving communication signals for halting generation/disconnecting from the
utility supply, if the utility requires such. Provision for required integration with SCADA or
other system should be made available for the utility at no charge.
• Generators larger than 100 kVA or generator units connected to a single POC whose
aggregated generation capacity exceeds 100 kVA should be controllable, i.e. active output
power should be controlled based on network conditions.

3.1.2.2 Safety protection and control


Requirements for normal operation and power quality are given in the summary tables (

Report for
Doc No: Rev Page 46 of 88
Table 3 to Table 8) in section 3.1.2.1.1.
The clauses for safety protection and control, which are part of section 4.2 of the NRS 097-2-1
document, detail requirements for safe operation of embedded generators under statutory power
quality requirements, and contingencies for abnormal or out-of-bounds condition events. Safe
operation of embedded generators considers both people and equipment, i.e. [24]:
• People safety, which includes:
o Owner (including personnel and/or inhabitants of the property) of the embedded
generator
o General public safety
o Utility personnel
o General emergency response personnel.
• Equipment safety, which includes:
o Utility equipment
o Other customers’ equipment connected to the same network(s)
o Generator equipment.
Safety protection and control requirements should be in accordance with SANS/IEC 62109-1 and
IEC 62109-2 [24]. Topics covered under safety protection and control include disconnection devices,
out-of-bounds voltage and frequency conditions, prevention of islanding, response to utility recovery,
isolation of embedded generators from utility, earthing, short-circuit protection and labelling. For out-
of-bounds conditions, embedded generators are required to trip off the utility network in accordance
to the response schemes outlined in Table 10. Some of the details in this table can also be found in

Report for
Doc No: Rev Page 47 of 88
Table 3 and Table 4. The responses to abnormal conditions listed in Table 10 apply to generators
under sub-categories A1 and A2. For generators under A3, disconnection should be in accordance
with the RPP Grid Code [24].
Table 10: Response to abnormal voltage and frequency conditions for SSEG in sub-categories A1 and
A2 [19]

Voltage range (at POC) Maximum trip time (in seconds)


V < 50% 0.2
50% ≤ V < 85% 10
85% ≤ V ≤ 110% Continuous operation
110% < V < 115% 40
115% ≤ V < 120% 2
120% ≤ V 0.16
Frequency range (at POC) Maximum trip time (in seconds)
𝒇 < 47 Hz 0.2
47 Hz ≤ 𝒇 < 50.5 Hz Continuous operation
50.5 Hz ≤ 𝒇 < 52 Hz Liner reduction of output power based on Figure 10.
52 Hz ≤ 𝒇 0.5

3.1.2.3 Metering
The NRS 097-2-1 makes reference to metering standards found in SANS 474/NRS 057 and
SANS 473/NRS 071. The metering guidelines cover three configurations that are determined by the
tariff structure required or implemented by the energy supplier. The type of metering is determined
by the utility based on the application and location of the embedded generator [24]. Figure 11 and
Figure 12 show the different configurations for single and multiple meter installations, respectively.
Details for single and multiple meter installations are compared in Table 11.

Report for
Doc No: Rev Page 48 of 88
Figure 11: Single meter installation [24]

Figure 12: Multiple meter installation separate metering (left) and


separate embedded metering (right) [24]

Report for
Doc No: Rev Page 49 of 88
Table 11: Comparison of single and multiple meter installations

Metering Configurations

Single meter installation Multiple meter installation

➢ Embedded generation feeds into the ➢ Embedded generation not directly


customer network (L) connected to customer’s distribution board

➢ Overall consumption and generation of ➢ Overall consumption and overall generation


customer is not recorded recorded
➢ Net energy import and export is metered ➢ Overall consumption balanced off between
and balanced embedded generation meter and net meter
(separated embedded metering
configuration)

➢ Net meter records and balances energy in ➢ Feed-in-tariff (FIT) metering records energy
a single register/bi-directional meter records generated from embedded generation and
energy import and export in separate reimburses customer at set FIT, while
registers, later balanced for billing system embedded generation consumption is
recorded and billed through conventional
means. Customer consumption billed
separately.

To achieve metering requirements, the types of meters used with embedded generation should
adhere to certain conditions. Pre-payment meters are not covered as they currently do not cater for
embedded generation. The following is a list of conditions for meter types:
• Energy meters used with embedded generation shall record active energy.
• Four-quadrant electronic meters shall be used in cases where embedded generation
installation require active and reactive energy recordings.
• Meters capable of metering quality of supply parameters shall activate the monitoring facility
on the meter.

3.1.3 The NRS 097-2-3: Simplified utility connection criteria for LV-
connected generators
It is not practically feasible for distribution network planners and designers to perform detailed
network studies for every SSEG application and potential connection to the utility network. As a
result, the NRS 091-2-3 provides guidance to South African distribution licensees for performing
simple assessment for SSEG connection to the utility. The NRS 097-2-3 proposes a set of criteria
based on international best practice that indicate the conditions under which embedded generators
can be connected to the LV network without performing detailed network studies [20]. These

Report for
Doc No: Rev Page 50 of 88
proposed criteria ensure that technical issues are minimised to allow customers to connect
generators without the need for mobilising resources to perform data intensive detailed analysis [22].
Similar to the NRS 097-2-1, the NRS 097-2-3 outlines a Scope, Normative references, Terms and
definitions, and the Requirements. The Scope of the NRS 097-2-3 covers the requirements for
simplified utility connection embedded generators for customers supplied by either shared or
dedicated LV networks, but explicitly exclude lower income domestic electrification networks, i.e.
shared LV networks supplying customers with a Living Standard Measure below seven [20].
The requirements outlined in the NRS 097-2-3 are categorised into general, requirements for shared
LV feeders, and requirements for dedicated LV feeders.
These are discussed in sections 3.1.3.1, 3.1.3.2, and 3.1.3.3. Figure 13 shows a summary of the
simplified connection criteria, and conveys the process for evaluating a generator for simplified
connection. Figure 14 shows the process involved in performing the simplified technical evaluation
for SSEG connection.

Figure 13: Summary of simplified connection criteria [20]


3.1.3.1 General Requirements under NRS 097-2-3
The general requirements for the proposed simplified connection criteria include rules that are
subject to the network design and the continuous capturing and documenting of details of all
generating plants, which include generator/plant size, type of generation, location, and date of
installation [20]. These requirements also apply in cases where the network feeder design After

Report for
Doc No: Rev Page 51 of 88
Diversity Maximum Demand (ADMD) is unknown. Some of the important general simplified
connection criteria are as follows [20]:
• Customers are limited to installing generation capacity size of up to 25% of the NMD. This
allows for 30% to 50% penetration levels on the network.
• Simplified connection of generator size should be limited to 350 kVA.
• Fault levels at the customer point of supply should be greater than 210 A.
• The total generation (shared LV generation and dedicated LV generation) supplied by a
MV/LV transformer should not exceed 75% of the MV/LV transformer rating.
• The total generation supplied by a MV feeder should not exceed 15% of the MV feeder peak
load.

3.1.3.2 Shared LV Feeders


Shared network feeders are sections of the utility network that supply more than one customer or
generator, as shown in Figure 13 [20]. Some of the important criteria for shared network feeders are
as follows [20]:
• The maximum individual generation limit is 25% of the customer’s NMD, up to a maximum of
20 kVA. Capacities greater than 20 kVA should be connected through a dedicated LV feeder.
• Generators greater than 4.6 kVA should be balanced across all phases.
• Alternative to being connected to a dedicated feeder, customers with generation capacities
exceeding 25% of the NMD limit can apply for an increased NMD.
• The total generation supplied by a shared LV feeder should be limited to 25% of the MV/LV
transformer rating.

3.1.3.3 Dedicated LV Feeders


In contrast to the shared LV feeder, the dedicated feeder exclusively supplies a single
customer/generator as shown in Figure 13. This allows for higher limits as compared to those
proposed for the shared LV feeder. The following are some of the important recommended criteria
for connection of embedded generation to a dedicated feeder [20]:
• The maximum generator size is limited to 75% of the NMD. Generators greater than 4.6 kVA
should be balanced across all available phases. In addition to being limited to 75% of NMD,
single phase connections should not exceed 13.8 kVA.
• Generator size should be limited such that the voltage change across the feeder (i.e. between
point of supply and transformer busbar) should not exceed 1% of the nominal voltage when
the generator transitions from zero to maximum generation, or from maximum to zero (e.g.
when the generator trips). Several graphs and look-up tables are featured in the NRS 097-2-
3 that provide generation capacity based on feeder cable parameters.
• Customers that only connect generators and have no load will be connected through a
dedicated feeder.

Report for
Doc No: Rev Page 52 of 88
Figure 14: Flow chart of simplified connection technical evaluation criteria [20]

Report for
Doc No: Rev Page 53 of 88
3.2 MPE guideline
The MPE guideline, developed by Moeller and Poeller Engineering in 2018, is a recommended
practice that can be used to assess connection of small generators connected to LV and MV grids
using hand calculations only. The guideline covers SSEG integration with capacities < 100 kW and
100 kW to 5 MVA. Simplified methods for the assessment of SSEG integration are covered and it
outlines conditions where more detailed analysis methods are required for potential SSEG
integration. It also provides guidance to determine the impact of connecting generation to the MV
grid. It adheres to requirements according to NRS 048, which is a standard on voltage compatibility:
• Supply voltage limits:
o MV networks: > 500 V, acceptable voltage limits are ± 5%
o LV networks: < 500 V, acceptable voltage limits are ± 10%
• Compatibility level for voltage unbalance on LV, MV, HV three-phase networks is 2%,
although on an LV network 3% may be applied.
The guideline contains design rules for LV networks (shared networks and dedicated networks) and
MV networks with the purpose of limiting voltage rises across a network when SSEG is to be
integrated. The simplified analysis method for assessing LV generator connections in the MPE
guideline is summarised in Figure 15. A brief description of the processes in the simplified analysis
method are:
1. Data collection: Data of LV generator that is to be connected (technical data), network data
and records of existing generation.
2. Calculate hosting capacity of service connection feeder: The maximum allowable generation
connected that can safely be hosted on the service connection feeder connecting the
generator to the network. The guide outlines hand calculations that can be used or look-up
tables with limited hosting capacities that have been determined based on typical LV cable
parameters.
3. Calculate hosting capacity of affected feeder: The maximum allowable generation connected
that can safely be hosted on the affected LV feeder (this is from the terminal bus or POC to
the MV/LV transformer in a network).
4. Calculate hosting capacity of LV network: The maximum allowable generation connected that
can safely be hosted on the LV network.
5. Determine possibility of connection: Using the flow chart in Figure 16.
6. Calculate maximum MV voltage allowable: Calculating maximum allowed MV voltage that is
dependent on allocated voltage rise in LV network, transformer nominal voltage ratio, and
transformer tap settings.
7. Documentation of assessment: Consists of a short report on each grid study undertaken and
included calculations and sources of data for the assessment carried out.

Report for
Doc No: Rev Page 54 of 88
Network strengthening Perform detailed analysis Network strengthening
required and re-evaluate required and re-evaluate
connection connection

No No No

SSEG rated Maximum Yes Maximum


power < Yes simultaneous simultaneous
Start service power of all power of gen in
connection gen. in affected LV network <
HC? feeder < HC of gen. capacity of
feeder? LV network?

Network strengthening
Yes required and increase
HC

Maximum

Approve SSEG Yes simultaneous power No


of gen in LV
connection network in upstream
MV network < HC of
MV network?

Figure 15: Simplified analysis method

Report for
Doc No: Rev Page 55 of 88
2. Calculate hosting
1. Data 3. Calculate 4. Calculate
capacity of service
collection hosting overall hosting
connection feeder
capacity of capacity of LV
affected feeder network

7. Documentation of 5. Determine
6. Calculate maximum
assessment possibility of
MV voltage allowable
connection

Figure 16: Determining possibility of connection of LV generator

3.3 Municipal training on simplified and detailed connection criteria

3.3.1 Simplified connection criteria: NRS 097-2-3 and MPE guideline


To offer capacity-building support to municipalities, the CSIR provided online training on SSEG grid
integration. The training aimed to provide guidance in addressing the need for skills by highlighting
simplified methods and detailed methods available for assessing SSEG integration and how these
can be used. The focus of the training included an introduction to SSEG grid impacts, simplified
analysis methods for SSEG integration, namely the NRS 097-2-3 simplified connection criteria and
MPE guideline. It also included detailed assessment methods for SSEG interconnection. The training
consisted of information sharing primarily and was then supplemented by interactive online quizzes
and worked examples. Due to its brief duration of two days, the training course aimed to provide a
basic understanding in SSEG integration methods. It included covering the theory of the different

Report for
Doc No: Rev
sections of the content followed by quiz sessions on the content and a general Q&A session.
Demonstration of the application of the content of various sections was included as were examples
for trainees to apply content.

A future training session for various aspects in grid impact studies is planned as there is ongoing
interest by municipalities in this regard.

3.3.2 Detailed connection criteria: Detailed grid impact studies


Detailed interconnection studies are required to confirm if voltages exceed acceptable limits and if
any equipment will be overloaded, in cases where simplified connection criteria fall short of
assessing a possible connection.

The following information under maximum and minimum loading and generation must be known:

• Acceptable voltage limits


• Thermal ratings on network equipment.

Detailed interconnection studies also consider voltage changes of the combined impact of all SSEGs
on the network. These studies include a fault-level check to ensure that equipment short-circuit
ratings are not exceeded and feeder protection is not compromised. Table 12 tabulates
recommendations on conducting detailed interconnection studies for LV and MV networks.
Table 12: Recommendations on when to conduct detailed interconnection studies

RECOMMENDATIONS TO CONDUCT DETAILED INTERCONNECTION STUDIES

LV NETWORK MV NETWORK
➢ When the simplified connection criteria as per the ➢ Verifying the feasibility of a new
NRS 097-2-3 is violated/for generators greater than generator connection at MV level.
350 kVA when the hosting capacity assessment
method is not applied as per the MPE guideline.

➢ One or more generators are to be operated at non- ➢ Further developing the MV network
unity power factor. to increase the hosting capacity.

➢ MV/LV transformer rating is more than 600 kVA.

➢ There is a high degree of unbalance anticipated.

➢ Customer reports voltage problems.


Note: It is important for the municipality to specify where the concern is and why there is a concern
(is it that the voltage rises, the fault level etc.). However, it is recommended that all aspects
of concern are considered.

Report for
Doc No: Rev
CSIR Simplified load flow tool – overview
The CSIR developed an Excel-based load flow tool to assist with addressing the lack of tools for
assessing SSEG integration. Its intended use is performing simple three-phase balanced load flow
assessments on radial distribution power networks to determine the steady-state operating
conditions with a given loading and/or be used to determine the effect on load flow of the addition of
SSEG to down-stream power networks (which includes small network equivalents of aggregated,
larger networks) of specific configurations of up to four bus bars. Microsoft Excel is the platform of
choice as it is widely accessible and has a relative minimal learning curve.

Of several methods of load flow calculation considered, i.e. Newton-Raphson, Gauss-Seidel and
Forward-Backward method based on Kirchhoff’s laws, the Forward-Backward method is deemed
the most suitable for the load flow calculation for simple radial distribution networks. The simplified
load flow tool is initially set up to compute load flow calculation, which includes finding the steady-
state voltages, power flows, thermal/voltage limit information for up to four bus bar networks with
different combinations. The combinations consist of a limited number of distribution lines and
transformers, and for the configurations it is set up for, has shown similarity to the commercial tool
DigSILENT PowerFactory for two, three and four bus systems of 99.9%, 99.8% and 97.9%,
respectively. For the municipal networks, the highly complex power systems can be reduced to their
aggregated equivalent networks. The load flow for these networks can then be computed by the load
flow tool to determine the effects of SSEG added at specific points in a network.

Figure 17: Homepage of CSIR simplified load flow tool

Section 4 covers the CSIR Simplified Load flow tool in detail.


(Next section: CSIR hosting capacity analysis tool – overview)

Report for
Doc No: Rev
4 CSIR Simplified load flow assessment tool
The CSIR, in partnership with various municipalities, assessed technical grid impacts of SSEG on
their networks to assess their readiness for SSEG integration as detailed in section 2. The CSIR
developed a simplified Excel-based load flow assessment tool as part of municipal support to
address the metric of availability and application of tools appropriate for assessing SSEG integration.
It can be used in place of manual hand calculations for specific network configurations.
Load flow or power flow is the process of finding a solution for steady-state quantities that is used to
gain insight into system performance over a range of operating conditions and are commonly used
to evaluate real and reactive power flow in the branches of a power network. Iterative methods are
used to obtain the solution, as the problem is non-linear.

Output:
- Voltage
Input: magnitude
Network data Load flow and angle
(Bus data, line calculation - Real/reactive
data, generator method power flow
data, load data) - Current flows
- Violations

Figure 18: Load flow process

4.1 Tool description

The simplified load flow tool’s main purpose is to perform simple load flow on relatively small
networks (including small network equivalents of aggregated, larger networks) to determine what the
effects are of desired SSEG addition to the network. The branch elements considered in the tool are
distribution lines and transformers. Table 13 details a summary of the properties of the load flow tool
including the capabilities, limitations, features and assumptions.
Figure 18 shows the load flow method that the tool uses. For a load flow calculation, the minimum
required data is to start by specifying the system power base, the configuration for which the
calculation will be performed, nominal voltages, transformer data/conductor data (where applicable)
and load/SSEG data. Once this information has been entered into the tool, a load flow method called
the Forward-Backward method, based on Kirchhoff’s Laws for voltage and current, is used to
compute load flow. This method was chosen compared to other conventional, widely used load flow
methods such as Newton-Raphson and Gauss-Seidel methods. Conventional load flow methods
tend to be ill-conditioned for distribution networks [28] whereas the Forward-Backward method is a
suitable method for radial networks. A summary of this method is found in Figure 18.

Report for
Doc No: Rev
Table 13: Tool properties

Tool capabilities Tool limitations Tool features Assumptions

➢ Three-phase ➢ Size limitations of radial power systems that the CSIR load flow tool ➢ The following prompts: ➢ Networks
balanced load can perform load flow calculation for: entered in
Load flow solution is acceptable.
flow calculation the tool are
(Forward- Parameter Two bus Three bus Four bus feasible and
Backward systems systems systems their have a
method based convergent
on Kirchhoff’s Number of distribution Min: 0 Min: 0 Min: 1 solution.
Law) lines
Max: 2 Max: 2 Max: 4 ➢ Generation
➢ Applicable to and load
Number of transformers Min: 0 Min: 0 Min: 0
radial networks Load flow solution does not converge. capacities
Max: 2 Max: 2 Max: 2 are known.
➢ View summary
of input data Maximum number of 2 2 2 ➢ Initial
branches in parallel voltages of
➢ View load flow
all bus bars
results Number of branches Min: 1 Min: 2 Min: 3 assumed to
➢ View individual Max: 2 Max: 4 Max: 6 be 1 p.u.
connected
➢ At most two bus bars and at most four bus bars Chosen limits for violations have been surpassed ➢ For p.u.
load/s and
(±10% LV, ± 5% MV) − both thermal and voltage conversion,
SSEG details ➢ Number of configurations that the load flow tool can compute load
violations. Vbase = Vnominal
and which bus flow for: and Sbase =
bar/s they are
o Two bus systems: 4 configurations 100 kVA or
connected to.
100 MVA
o Three bus systems: 8 configurations (either can
o Four bus systems: 34 configurations be selected)

Report for
Doc No: Rev Page 60 of 88
Required input data
• System power base [kVA or MVA]
• Configuration details:
o Number of buses
o Number of distribution lines
o Number of transformers
• Nominal voltages1 [V or kV]
• Transformer data2
• Conductor data2
• Load data (Capacity [kVA or MVA]; Power factor (p.u.))
• SSEG data2 (Capacity [kVA or MVA]; Power factor (p.u.))

Load flow calculation: Forward-backward sweep method

Known:
Apparent power of
loads/generation
System
impedances

Mismatch between
successive iterations
is < 𝟏𝟎−𝟑 in p.u. form,
i.e. values obtained
are steady.

Load flow results


• Steady-state bus bar voltages
• Real and reactive power flows
• Thermal violation information (voltage violations and thermal violations)
Report for
Doc No: Rev Page 61 of 88
Figure 19: Load flow method
1Nominal voltages must be specified by the user if no transformers are present in the network for which load flow is
calculated. In the case of no transformers present in the system (network branches consist of only distribution lines), the
nominal voltage selected for ‘bus 1’, which is chosen as the bus bar closest to the main generation source, is then assumed
for all the other bus bars.
2 This is required if this network element is present in the desired configuration. [29]

Load SSEG

Figure 20: Single line diagram of example of radial three bus network configuration in load flow tool
(Two parallel distribution lines between bus 1-2 and two parallel transformers between bus 2-3)

4.2 Tool validation


4.2.1 Configuration testing
To verify the accuracy of the load flow tool (particularly comparing the steady-state bus voltage
computation and power flows), the configurations for which it is set up to compute load flow, were
tested against the commercial tool DigSILENT PowerFactory.
Table 14 is a summary of the similarity between the tool and the DigSILENT PowerFactory. Test
results indicate that the tool is reliable for computing various load flow calculations for a range of
configurations. For the four bus systems it must be noted that the percentage similarity for power
flows of the load flow tool to DigSILENT is slightly lower than two or three bus systems, at 96.2%.
This could probably be attributed to rounding differences between the different software or the
different load flow methods used.

Report for
Doc No: Rev Page 62 of 88
Table 14: Percentage similarity between load flow tool and DigSILENT PowerFactory

Size of system Average percentage similarity between load flow Overall percentage
tool and DigSILENT PowerFactory similarity

Bus voltages Power flows

Two bus systems 99.75% 99.75% 99.75%


(four
configurations)

Three bus 100% 99.9% 99.9%


systems (eight
configurations)

Four bus systems 99.5% 96.2% 97.9%


(32
configurations)

4.2.2 Municipal test cases


Since the main intended users are various municipalities, municipal network test cases are used on
the load flow tool. The networks are larger than the tool’s configurations, hence aggregation is
necessary for use on the load flow tool, which will allow one to observe the overall effect of adding
SSEG/loads to the down-stream networks. The aggregated networks attempt to mimic the overall
network as closely as possible, but it is important to note for comparison purposes that the load flow
tool results match the aggregated network modelled on DigSILENT more closely as this would have
the exact same configuration. The test cases are shown in Figure 21 and Figure 22.

Report for
Doc No: Rev Page 63 of 88
Figure 21: Test case 1: 900 kVA SSEG application (Municipal test case 1 − Three bus system)

Figure 22: Test case 2: 560 kVA SSEG application (Municipal test case 2 – Four bus
system)

Report for
Doc No: Rev Page 64 of 88
Table 15 is a summary of the results from the municipal cases. Each test case analysed four load
flow scenarios, which included running a load flow computation when there is maximum loading and
no SSEG in the system, maximum loading with the SSEG connected, minimum loading with no
SSEG and minimum loading with SSEG connected. The reliability of the computation capability of
the load flow tool is again shown here.

Table 15: Results for municipal test cases

Test case Average percentage similarity between load Overall percentage


flow tool and DigSILENT PowerFactory similarity
Bus voltages Power flows
Test case 1 (900 kVA 99.6% 99.1% 99.4%
SSEG application)
Test case 2 (560 kVA 99.7% 99.7% 99.7%
kVA)

4.3 Key outcomes and learnings

• The overall percentages of similarities to DigSILENT PowerFactory for the different tests on
the load flow tool for two, three and four bus configurations were 99.9%, 97.5% and 97.9%,
respectively. For the municipal test cases (aggregated networks), percentage similarities of
99.4% and 99.7% for test case 1 and test case 2 were shown – an indication that the load
flow tool is reliable for various simple load flow computations. (The main comparisons being
steady-state voltage and power-flow calculations.)
• The Forward-Backward method, although relatively simple to implement on Excel/Excel VBA
and relying on linear back/forward sweep equations based on Kirchhoff’s laws, becomes
unsuitable for closed loop networks. Computation becomes intensive if the configurations
were to increase in the number of bus bars or network configuration complexity in the system.
Excel used purely as the computation engine is limited with the load flow method used, even
though it is a highly accessible, relatively cost-effective platform with a minimal learning
curve.
• From user evaluations from various municipalities:
• There were some comments around the ease of use since it is the first version of this
tool.
• The tool allows a user to input cable information from a datasheet – a suggestion was
put forward that having standard cable sizes built-in would streamline the load flow
process and possibly reduce input errors.
• A suggestion to allow a user to specify the initial voltage at various bus bars as one
would maybe prefer to conduct a load flow when the initial voltage is some value other
than the nominal.

Report for
Doc No: Rev Page 65 of 88
4.4 Recommendations and future development
• If the tool is to be developed further, an alternative platform than Excel/Excel VBA may be
required to develop the computation engine for load flow as the computational power is
limited.
• Although the focus of testing with municipal cases at this stage was to verify the accuracy of
the load flow tool primarily by using aggregated models of the municipal networks, which
were not exact replicas of the actual models, more accurate modelling of the municipal
networks when aggregating will be necessary, which will improve confidence in the load flow
tool’s results.
• The tool requires benchmarking with other Excel-based or similar tools that have comparable
functionality to determine what future modifications may be needed to set it apart from
existing software/programs.

Report for
Doc No: Rev Page 66 of 88
CSIR hosting capacity analysis tool – overview
Some parts of South Africa’s metropolitan municipalities show an increasing rate in the deployment
of Small-Scale Embedded Generation (SSEG). SSEG predominantly consist of photovoltaic (PV)
modules and due to the variability of solar energy, the output power does not always match the load
profile. This can cause challenges to the distribution grid network and can sometimes violate the
defined network operational limits.
Network planners and operators can plan better for system upgrades when feeder hosting capacity
limits are calculated. A feeder hosting capacity analysis tool is developed to enable South African
distribution planners to determine the thresholds at which SSEG can be integrated into their
distribution networks without undertaking network strengthening, and to determine locations for
network improvements to allow for additional uptake of embedded generation capacity beyond this
threshold.
The CSIR developed a hosting capacity analysis tool using two methods, viz. deterministic and
stochastic. The two methods were implemented in DigSILENT PowerFactory using DigSILENT
Programming Language (DPL). These methods were assessed to determine their advantages and
disadvantages. The methods were tested using a South African municipal test network during which
results were compared and used to assess functionality, capability, and to develop recommendations
for improvements. Table 16 gives a comparison of the two hosting capacity methods tested.
Table 16: Comparison of deterministic and stochastic hosting capacity methods

DETERMINISTIC HOSTING CAPACITY METHOD STOCHASTIC HOSTING CAPACITY METHOD

Faster computational time Longer computational time

Simple to implement Implementation and interpretation of results is more


involved

Does not reflect reality as it uses a fixed capacity and Simulates reality as analysis accounts for
assesses a single node at a time randomness in generator locations and sizes

High precision at each node as multiple evaluations Low precision at each node as multiple evaluations
yield the exact same results may not yield the exact same results

Despite having a longer computation time, the stochastic hosting capacity method accounts for the
probabilistic behaviour of SSEG installations observed in municipality networks, which yields results
that are more accurate and useful to system operators and planners. Thus, the stochastic method
would be the method more appropriate for calculating feeder hosting capacity.
Section 5 covers the hosting capacity analysis tool in more detail.
(Next section: Recommendations on NRS 097 – overview)

Report for
Doc No: Rev Page 67 of 88
5 CSIR hosting capacity analysis tool
There is a growing concern for the safety, reliability and quality of supply, which requires a proactive
response from distribution network planners and operators. Several methods of analysis exist that
can be implemented for interconnection and planning studies. One such a method is performing an
analysis that assesses the network's ability to host more SSEGs, known as Hosting Capacity
Analysis.

Hosting capacity is the amount of generating power (SSEG) that can be added to the network without
violating the network integrity. Control system upgrades are required to safely integrate additional
generation (SSEG) [30][31]. Hosting Capacity Analysis can help municipalities understand the
impact of adding new SSEGs to the electrical distribution system and further provide more
information on the associated costs needed to upgrade the distribution network to accommodate
more SSEGs to the network. Hosting Capacity Analysis provides information on the potential system
impacts of SSEG when generation is increased, different interconnection options, and the trade-offs
between cost and hosting capacity expansion for a range of possible distribution system upgrades
that could be used to integrate SSEG. The CSIR team developed a hosting capacity tool for South
African municipalities to determine the potential impacts that may occur when increasing SSEG
interconnections to the distribution network. This tool is developed using DigSILENT PowerFactory
software and can only be used with this specific software.

Most methods use a similar principle as seen in Figure 23, whereby SSEG penetration is increased
in user-defined step sizes at a location while performance indices are checked at every iteration until
violation occurs, then hosting capacity is obtained. All feeders will have a unique response to the
interconnection of SSEG and therefore other impact factors of calculating hosting capacity can be
the feeder, PV deployments and specific utility established thresholds [32]. To develop a hosting
capacity tool for South African municipalities, two methods were considered, namely the
deterministic hosting capacity and the stochastic hosting capacity method.

Report for
Doc No: Rev Page 68 of 88
Figure 23: Typical flow chart to calculate hosting capacity

5.1.1 Description of the developed Deterministic Hosting Capacity Tool


Figure 24 shows the process for calculating the hosting capacity using the deterministic method. To
perform calculations, the feeder circuit model, maximum and minimum load data and equipment
ratings are required. A single busbar location is selected for PV system installation and the size of
the generator is increased by a constant value. To select the constant value for increasing
generation, it is recommended that a study is done on the behaviour of customers in the area in
question [31]. The typical PV system size for residential areas is 5 kW [33]. Impact factors (voltages
and thermal overloading) are evaluated using a load flow calculation method and over-voltage,
under-voltage, cable limitation and transformer limitations are evaluated for any violations. The
violation limits set in the code are as follows:
• For voltages less than 500 V, acceptable voltage deviation is ±10%
• For voltages greater than 500 V, acceptable voltage deviation is ±5%
• Line and transformer thermal loading is ≤ 100% of continuous thermal rating.
Voltage limits are from the requirements given in the NRS 048-2 [26]. The process is repeated for
pre-selected locations to determine the hosting capacity of the nodes.

Report for
Doc No: Rev Page 69 of 88
Figure 24: Flow chart for the Deterministic Hosting Capacity Method

5.1.2 Description of the developed Stochastic Hosting Capacity Tool


The stochastic hosting capacity analysis method implemented in DPL scripting uses a random
number generator1 to introduce randomness in bus selection and SSEG capacity increase. The type
of random number generator used is the Mersenne Twister, a type of random number generator
classified as a strong pseudo-random number generator [34]. This means that this type of random
number generator has a long period (number of random values generated before repeating a
sequence) and uniform distribution of values [35].
Figure 255 shows a flow diagram that illustrates the algorithm used to carry out the stochastic hosting
capacity analysis. The algorithm starts much the same way as the deterministic method, in that a

Report for
Doc No: Rev Page 70 of 88
network model is created or chosen and the characteristic data of the network, loads and generators
are loaded onto the model. The algorithm then runs a load flow test, and determines the voltage and
thermal limits of the network components.
The algorithm continues by selecting the bus corresponding to a distributed generator at random,
using the random number generator setup. If the algorithm is on its second, third or later iteration, it
tests whether the bus selected is the same as that in the previous iteration. If so, a reselection takes
place, otherwise the algorithm continues to generate a random size with which to increase the
generator capacity, also using the same random number generator method. The capacity size range
is between 50 and 100 KW (selected arbitrarily).
After increasing the capacity on the selected bus, a load flow test is executed. The algorithm then
checks for any voltage violations on buses or thermal violations on lines and transformers. The
violation limits are the same as those for the deterministic method described in section 5.1.1. If there
are no violations detected, the algorithm loops back to randomly selecting another bus and runs
through this process again. If a violation is detected, the algorithm subtracts the recently added
capacity and records the current active power capacity on the bus as the hosting capacity of that
bus. Once all buses of interest have been accounted for, the algorithm generates a list of hosting
capacities.

Figure 25: Stochastic hosting capacity method flow chart

Report for
Doc No: Rev Page 71 of 88
5.2 Tool comparison of developed methods
It is worth noting that several hosting capacity methods exist. For assessments on the hosting
capacity tool developed, only the two methods described are examined and compared against each
other for capability and functionality. Accuracy of the methods can only be evaluated when all impact
factors are considered because impact factors determine hosting capacity, therefore the more impact
factors are evaluated the more accurate is the hosting capacity calculation. Method comparison for
the study looked at was the hosting capacity calculated in terms of the method’s ability to reflect
reality and computational time.
Table 17: Comparison of the deterministic and stochastic hosting capacity methods

DETERMINISTIC HOSTING CAPACITY METHOD STOCHASTIC HOSTING CAPACITY METHOD

Faster computational time Longer computational time

Simple to implement Implementation and interpretation of results are


more involved

Does not reflect reality as it uses a fixed capacity and Simulates reality as analysis accounts for
assesses a single node at a time randomness in generator locations and sizes

High precision at each node as multiple evaluations Low precision at each node as multiple evaluations
yield the exact same results may not yield the exact same results

5.2.1 Hosting capacity output results for two terminals


The hosting capacity analysis was undertaken for the stochastic method and compared to the
deterministic method to observe functionality. Results were assessed for a South African
metropolitan municipal feeder network. Hosting capacity was observed for two terminals/buses
(hereafter referred to as node 4 and node 5) that had the largest hosting capacity values. At base
case the hosting capacity values are approximately the same for both nodes when calculated using
the deterministic and stochastic methods. The values begin to change when increasing the number
of nodes assessed for only the stochastic method. The trends (linear graphs) at which the hosting

Report for
Doc No: Rev Page 72 of 88
capacity of nodes 4 and 5 decrease as a function of assessed nodes for the stochastic method are
shown in Figure 266. The points represent the actual values obtained during testing.

Figure 26: Hosting capacity calculated on two different terminals using both the deterministic and
stochastic method

5.2.2 Computational time comparison


The computational times for the stochastic and deterministic methods are compared in Figure Figure
277. The figure shows the trends (linear graphs) of increasing computational time for the two
methods as a function of nodes assessed. Again, the points represent the actual values obtained
during testing. Figure 277 shows that the computational time for the stochastic method increases
faster than that of the deterministic method as nodes assessed increase. This observation is a result
of the stochastic nature of the algorithm and the fact that the stochastic method considers generation
capacity at all nodes of interest during a single analysis, as opposed to the deterministic method that
only considers generation capacity at a single node. However, given the advantages that the
stochastic method provides (discussed in section 5.2.3), the added computational time can be
considered an acceptable trade-off.

Report for
Doc No: Rev Page 73 of 88
Figure 27: Computational time deterministic method vs stochastic method

5.2.3 Analysis of results


The deterministic hosting capacity method computes the hosting capacity of a single node at a time
under the assumption that no generation capacity is installed at any other node. The result is a
hosting capacity for a node that remains constant regardless of the number of nodes considered for
SSEG installation as seen in Figure 266. This method caters for the highly unlikely case of generation
being connected at a single point on the network. In reality, installation of SSEG happens at multiple
locations at random. As a result, the hosting capacity of a specific node could be reached due to the
threshold of a component connected elsewhere in the network and influenced by embedded
generation connected at multiple points on the network. The stochastic method accounts for this and
provides results based on the likelihood of generation being installed at multiple points in the
network. This is seen in Figure 266 where the hosting capacities of nodes 4 and 5 decrease with
increasing nodes assessed, showing that the hosting capacity of any node depends on generation
at other nodes. The stochastic method also provides more insight into crucial points in the network
that require upgrading to increase the overall hosting capacity.

A limitation with the stochastic method is its lack of precision. Observations of results from tests
conducted suggest the possibility of a weak correlation between hosting capacity outputs and
depicted trends of the stochastic method as seen in Figure 26. This suggests that the likelihood of
getting the exact same results from multiple tests is low. This is due to the random nature of the
algorithm employed. Additionally, the stochastic method is computationally intensive. From Figure
27 it can be seen that the stochastic method has a long computational time that increases at a higher
rate with increasing nodes as compared to the deterministic method. However, depending on the
intended use of the tool and the size and structure of network being studied, longer computational
times might not be an issue.

Report for
Doc No: Rev Page 74 of 88
5.3 Recommendations

The stochastic hosting capacity method is recommended to be applied when calculating hosting
capacity on South African municipal networks for the following reasons:
• The intended use of the tool is from the customer-based perspective because the main
objective identified is to plan for the increasing rate of SSEG applications in some parts of
South African metropolitans. The rate of integration has unknown variables such as the type
of SSEG installation, the size of the generator and where the generator will be installed; the
stochastic hosting capacity method allows for these uncertainties.
• The stochastic hosting capacity method allows for realistic modelling. As time goes by, a
trend will develop in some areas, and that can be further implemented on the stochastic
hosting capacity method to calculate an even more realistic hosting capacity of a feeder.
• An assumption is made that this tool will not be used every day but over certain periods of
time. From that assumption the issue related to computational time can be addressed.
However, in some cases the burden of time can be excessive, especially if the computer or
laptop cannot handle intensive computations. In such cases, networks can be simplified
(reduced) to a representation that consists of nodes of interest. This should ease the
computational burden of the hosting capacity analysis.

Report for
Doc No: Rev Page 75 of 88
Recommendations on NRS 097 – overview
The CSIR’s engagements with municipalities revealed some concerns that municipalities have
regarding the NRS specifications on SSEG integration into the distribution network. These concerns
include:
• The 350 kVA generator capacity limit for LV network under simplified connection criteria
outlined in the NRS 097-2-3.
• The lack of clear definitions and specifications on connections of uninterruptable power
supply (UPS) systems into the distribution network.
To address these concerns, the CSIR has undertaken an investigation into the issues to propose
potential adjustments and inclusion to the NRS specification.
• From the investigation, the basis for the 350 kVA limit was determined by two assumptions:
• The requirement that the total generation (i.e. shared LV generation and dedicated LV
generation) supplied by a MV/LV transformer should be less than 75% of the MV/LV
transformer rating.
• The highest typical rating of MV/LV transformers for South African LV distribution networks
is 500 kVA.
Additionally, the NRS specification limits its scope to SSEG systems that can potentially feedback
into the distribution network. This excludes most UPS system connections as they are typically
installed to prevent backflow, however these systems can still have adverse effects on the
distribution network.
The following aspects are recommendations for the NRS specification:
• Regarding the 350 kVA limit:
o The 350 kVA limit should be removed from the recommendations and more emphasis
should be placed on the maximum generation limit of 75% of MV/LV transformer
rating instead.
o Alternatively, an addendum should be made to the 350 kVA limit clause to state that:

“Simplified connection of generator sizes should be limited to 350 kVA. This is


based on the fact that the typical maximum MV/LV transformer rating for
distribution networks is 500 kVA with the 350 kVA limit informed by the
75% limit of MV/LV transformer rating. In instances where MV/LV
transformer rating is above 500 kVA, adhere to the 75% limit.”

• Regarding UPS connections, the following from the SANS standards can be incorporated:
o Section 4.6 of the SANS 62040-1
o Clauses 1.6.1, 1.6.2, and 1.6.4 of the SANS 62950-1
o The SANS 60947-6-1 and SANS 10142-1 for requirements on change-over switching
o In terms of additional switching devices and requirements, the SANS 60950-1 section
3.4 could be consulted [36]

Report for
Doc No: Rev Page 76 of 88
o Regarding islanding and backflow, the SANS 62040-1 clause 5.1.4 provides more
details on backfeed protection and prevention against hazardous islanding conditions
[36]. This clause could be included in islanding/backflow preotection requirements
specified in the NRS 097-2-1.

Section 6 covers the recommendations on NRS-097 in more detail.


(Next section: Resources )

Report for
Doc No: Rev Page 77 of 88
6 Recommendations on NRS-097
As mentioned in section 2, the CSIR evaluated several municipalities for their readiness to integrate
SSEG into their distribution networks. Evaluations focused on several aspects involving SSEG grid
impact assessments, which include business processes, personnel and skills capacity, availability
of data, and resource availability and adequacy. From engagements with municipalities during the
evaluation and gap analysis process, distribution licensees were concerned with the limit imposed
on SSEG connection size for LV networks, which specifies that connections are limited to a total
capacity of 350 kVA for LV networks [20]. This limitation suggests that customers who wish to install
capacities above 350 kVA will require detailed studies to be conducted at their own expense.
However, municipal distribution licensees are already installing capacities above this limitation
without performing detailed analyses and have yet to encounter adverse effects on their LV
networks. Hence, municipal distribution licensees are questioning this limitation, with the intention of
trying to ascertain the basis for this limit.
As an additional concern, the classification of connection types for electrical energy storage systems
(ESS), such as UPS systems, were also considered. Current grid codes and specifications lack clear
definitions on connections of such systems to distribution networks in terms of whether connections
can be considered as grid-tied or off-grid. This may result in customers who wish to install SSEG
with ESS as well as distribution licensees misinterpreting and disagreeing on how such systems
need to be integrated into distribution networks, in which case ESS systems that are technically grid-
tied may be classified and registered as off-grid. The misinterpretation could result in treatment of
grid-tied ESS as off-grid systems, which could have negative impacts on the network.

6.1 Objective
As a result, the CSIR investigated both matters – the issue of the 350 kVA limit, and the classification
of UPS systems – with the purpose of proposing recommendations for amendments to be made to
the NRS 097 documentation to address these concerns. The main objectives of the investigation
were:
• To address issues concerning the 350 kVA limit for MV connection applications and
determine what the basis of this limit is, whether this limit can be altered and to what extent.
• To clarify classification definitions for UPS systems and any other ESS system based on grid-
tied systems and off-grid systems.
• To make recommendations to the NRS 097 workgroup for amendments or inclusions to the
next iteration of the NRS 097 publication that will address these issues.

6.2 The 350 kVA Limitation for Simplified Connection Criteria


6.2.1 Basis of concern regarding the 350 KVA limit
The NRS 097-2-3 mentions that the limit for connections of generators into LV networks with
simplified connection criteria is 350 kVA. However, it is not made clear what the basis for this limit
is. Upon investigation, it was found that the limit is based on the following:

Report for
Doc No: Rev Page 78 of 88
a) The total generation (i.e. shared LV generation and dedicated LV generation) supplied by a
MV/LV transformer should be less than 75% of the MV/LV transformer rating [20].
b) The highest typical rating of MV/LV transformers for South African LV distribution networks
is 500 kVA.
The value of 350 kVA works out to be 70% of 500 kVA, which is a more conservative outcome of
applying the specification outlined in a).
While the 75% limit is well justified, applying this limit with the assumption that MV/LV transformers
for distribution networks can only be rated at a maximum of 500 kVA results in an SSEG integration
limitation that is too restrictive for networks that do in fact have MV/LV transformers that are rated
above 500 kVA. Such distribution networks can afford to integrate SSEG capacities above 350 kVA
without adverse effects on the network becoming a concern. As such, it is concluded that the
350 kVA limit warrants revision.

6.2.2 Basis of the capacity limit of 75% of transformer rating


The requirement for limiting the maximum aggregated generation capacity connected downstream
of an MV/LV transformer to 75% of the transformer rating is based on the voltage deviation limit
stated in the NRS 048-2 [26]. The 75% limit is also based on the rapid voltage change limit of 3% as
per recommended international standards such as the VDE AR-N-4105 German standard [20] [21].
While the calculations are not explicitly documented, the NRS 097-2-3 qualifies the 75% limit based
on the discussion and computation provided hereafter.
The combination of the 3% voltage rise limit and the ± 10% voltage range on the LV network is used
to define a design criteria for determining the maximum generation capacity to install on a LV network
[21]. Considering the 3% voltage rise limit recommended, it is proposed that for both shared and
dedicated feeder LV networks this 3% be divided as follows [21]:
(1) 1% voltage rise allocated to the MV/LV transformer.
(2) 1% voltage rise allocated to the feeder up to the last metering kiosk (three-phase cabling).
(3) 1% voltage rise allocated to the service connection feeder up to the point of connection
(often single-phase).

Condition (1) is crucial for the determination of the 75% limit. The maximum MV voltage for normal
operating conditions is 104% [20]. This value is based on the normal MV on load tap changer setting
[20]. The built-in boost for the standard MV/LV transformer in nominal tap is 5% [20]. From these
two conditions, the maximum LV voltage at the transformer under maximum MV voltage, no load
and no generation conditions is determined as follows:
𝑀𝑎𝑥𝑖𝑚𝑢𝑚 𝑀𝑉 𝑣𝑜𝑙𝑡𝑎𝑔𝑒 × 𝑡𝑟𝑎𝑛𝑠𝑓𝑜𝑟𝑚𝑒𝑟 𝑏𝑜𝑜𝑠𝑡 = 104% × 5% = 5.2%

𝑀𝑎𝑥𝑖𝑚𝑢𝑚 𝐿𝑉 𝑣𝑜𝑙𝑡𝑎𝑔𝑒 = 104% + 5.2% = 109.2% (1)

Typical impedance rating 𝑍 and 𝑋/𝑅 ratio for MV/LV transformers are 6% and 5.0, respectively [20].
From this, the transformer resistance can be calculated as follows:

Report for
Doc No: Rev Page 79 of 88
𝑍 = √𝑅 2 + 𝑋 2 = √𝑅 2 +(𝑋/𝑅 × 𝑅)2
6% = √𝑅 2 + (5 × 𝑅)2

∴ 𝑅 = 1.2% (2)

Considering the generation limit connected to the transformer, 𝑃𝑚𝑎𝑥 is capped at 75% of the
transformer rating, and an assumption is made that the generation is at unity power factor (i.e.
percentage voltage 𝑈 is 100%), the voltage rise across the transformer can be confirmed to be within
the 1% condition as follows:
𝑃𝑚𝑎𝑥 = 𝐼𝑚𝑎𝑥 × 𝑈
75% = 𝐼𝑚𝑎𝑥 × 100%

∴ 𝐼𝑚𝑎𝑥 = 75% (3)

The voltage rise ∆𝑈 across the transformer is calculated as:


∆𝑈 = 𝐼𝑚𝑎𝑥 × 𝑅
∆𝑈 = 75% × 1.2%

∴ ∆𝑈 = 0.9% (4)

Adding results from (1) and (4) gives the maximum transformer LV voltage at minimum load,
maximum generation conditions, calculated as:

109.2% + 0.9% = 110.1% (5)


The result in (5) shows that at a generation level of 75% of the MV/LV transformer rating and
minimum loading, the maximum LV voltage at the transformer’s LV terminal rises to 110.1% of the
nominal voltage. This is already slightly above the 110% limit proposed in the NRS 048-2
recommendations. If a further 1% rise occurs across the LV feeders, the LV voltage level would rise
to 111.1%. However, in reality there will be loading on the system that will reduce this voltage rise.
Thus, the generation limit of 75% is proposed, noting that certain conditions may result in the voltage
limit being exceeded and remedial action will be necessary to stabilise network conditions [20].

6.2.3 Recommended adjustment to 350 kVA limit requirement


The following suggested changes to the NRS specification is recommended for the next iteration of
the NRS 097-2-3 publication:

• The 350 kVA limit should be removed from the recommendations and more emphasis should
be placed on the maximum generation limit of 75% of MV/LV transformer rating instead.
• Alternatively, an addendum should be made to the 350 kVA limit clause to state that:
o Simplified connection of generator sizes should be limited to 350 kVA. This is based
on the fact that the typical maximum MV/LV transformer rating for distribution
networks is 500 kVA with the 350 kVA limit informed by the 75% limit of MV/LV

Report for
Doc No: Rev Page 80 of 88
transformer rating. In instances where the MV/LV transformer rating is above
500 kVA, adhere to the 75% limit.

In addition to the suggested changes to NRS 097-2-3, investigations into statistical methods and
analyses of typical maximum MV/LV transformer ratings for distributed networks should be
conducted to ascertain if the 350 kVA limit needs to be revised following the above suggested
recommendations. This could allow for increased hosting capacity of distribution networks in South
Africa and well-informed updated view on the existing 350 kVA limit.

6.3 Classification of UPS Systems


6.3.1 The Problem of SSEG with UPS Systems
The gap analysis into one of the municipal distribution licensees revealed a concern with the
classifications of SSEG given in “Requirements for Small-Scale Embedded Generation: Application
Process to become a small-scale embedded generator” [37].
The classifications given include the following:

Off-grid SSEG – An SSEG that is


Grid-tied SSEG – An SSEG that Grid-tied hybrid SSEG – Grid- physically separated and
is connected to the utility's tied SSEG that islands after electrically isolated from utility
electricity grid directly, through a interruption of the utility supply or electricity grid – either directly or
customer’s internal wiring, or when the applicable electrical through a customer’s internal
through a reverse power flow- service conditions are outside wiring. Consumer loads cannot be
blocking relay. stated limits or out of required simultaneously connected to the
tolerances and then supplies the utility grid and the SSEG
load from the inverter, operating in installation, and export of energy
the stored-energy mode via a into the utility grid by the generator
suitably interlocked change-over must not be possible. An SSEG
switch. that is connected to the grid
through a reverse power flow-
blocking relay is not considered to
be off-grid.

Report for
Doc No: Rev Page 81 of 88
In spite of these definitions, passive standby UPS systems with connection descriptions as outlined
in [18] – which makes provision for a physical connection to, and allows sourcing of power from the
utility grid – are classified as off-grid. Figure 288 shows the original classification scheme for
residential SSEG systems, with passive standby UPS systems classified as off-grid. Contention
between customers and distribution licensees may arise when the licensees try to impose restrictions
on the sizing of passive standby UPS systems installed by customers based on grid-tied principles,
as customers can contest these restrictions on the grounds that the systems are classified as off-
grid. Customers may thus install UPS systems with inverter sizes exceeding the After Diversity
Maximum Demand that they pay for, which may have adverse effects on the network as it is not
designed for such capacities.

Figure 28: Original classification of scheme for residential SSEG

Report for
Doc No: Rev Page 82 of 88
6.3.2 Recommended classification inclusion to the NRS documentation
The scope of discussion of UPS systems in the NRS documentation only includes a definition of the
term UPS, requirement specifications for parallel connections of UPS with the grid, and brief details
of earthing configurations of embedded generation with UPS. This information is sourced from
SANS 10142, which only specifies connection guidelines and output requirements of UPS systems,
however does not mention charging and sizing requirements [38]. Inverter specifications and
requirements outlined place emphasis on SSEG systems of the PV type. Due to this void of
information on UPS systems, incidences of misunderstanding and disagreements between
customers and distribution licensees may become more prevalent as the adoption of SSEG
applications increases. To avoid any such confusion or contention in future, it is recommended to
provide clear definitions and classifications of ESS systems, such as UPS, on future developed
SSEG application requirements in NRS documentation. Figure 299 provides a structure for the
recommended classification of systems (including UPS systems).

Figure 29: Proposed classification for residential SSEG

6.3.3 Recommended SANS standards to consult regarding UPS systems


Not all requirements for UPS systems are included in the NRS 097-2-1, and as a result, some of the
effects of possible abnormal operations of UPS systems may be overlooked, which could result in
problems in the distribution network, damage to equipment, or the injury to personnel responsible
for the distribution network or people utilising the UPS system. The IEC standards that speak to UPS
systems (IEC 62040) could be consulted to address some of these short-falls on UPS connections.
Since the IEC 62040 have been adopted as SANS 62040 standards, this means that they are
applicable in the South African context. The following discussion recommends several standards
that could be consulted for UPS connections that could be used to either make amendments to the
NRS 097 series or to be used in conjunction with the series. The discussion is part of an ongoing
investigation into practices, standards and regulations regarding UPS systems that could be used to
augment NRS requirements.

Report for
Doc No: Rev Page 83 of 88
Compliance of UPS connections, and any SSEG for that matter, to the NRS 097-2-1 is based on
whether there is a possibility of backflow of energy to the distribution network. Thus, UPS systems
that prevent backflow can connect to the utility without proper consideration of the effect charging
can have on the interface of the system with the distribution network. To address this gap, the
following could be consulted:

• Section 4.6 of SANS 62040-1 [39]


• Clauses 1.6.1, 1.6.2, and 1.6.4 of SANS 62950-1 [36].

These requirements speak to the charging currents for UPS systems. While the requirements are
based on the protection of the UPS systems and connected load, these aspects are worth
considering when accounting for the loading on the distribution network due to UPS charging.

The aspect of protection in the NRS 097-2-1 considers the disconnection switches and protection
schemes for protecting the distribution network. These requirements include discussions on the
disconnection switches, abnormal voltage and frequency protection, islanding, DC current injection,
and short-circuit protection [24]. The document even makes reference to SANS 10142-1 and the
SANS 60947-2 with regards to requirements preventing uncontrollable islanding [40]. However,
there is little mention of requirements of change-over switches and switching scheme. It is
recommended that the following be incorporated in the NRS 097-2-1:

• SANS 60947-6-1 and SANS 10142-1 for requirements on change-over switching.


• In terms of additional switching devices and requirements, SANS 60950-1 section 3.4 could
be consulted [36].
• Regarding islanding and backflow, the SANS 62040-1 clause 5.1.4 provides more details on
backfeed protection and prevention against hazardous islanding conditions [36]. This clause
could be included in islanding/backflow preotection requirements specified in the NRS 097-
2-1.

Report for
Doc No: Rev Page 84 of 88
7 Resources
Available resources are shown in Table 188, which may assist with SSEG integration information as mentioned
in this document:
Table 18: Resources that provide SSEG integration information

Resource Description

NRS 097 documentation NRS 097-2-1: Specification that covers requirements at the point of utility
connection for SSEG and prioritises safe interconnection of SSEG.(2017)
Available at: https://www.sseg.org.za/wp-content/uploads/2017/01/NRS-
097-2-1-2017-Edition-2.1-published-2020-07-20.pdf
NRS 097-2-3: Outlines the requirements for simplified utility connection
embedded generators for customers supplied by either shared or
dedicated LV networks. (2014)
Available at: https://www.sseg.org.za/wp-content/uploads/2019/03/NRS-
097-2-3-final-2014_B_W.pdf

The MPE guideline Recommended practice for assessing connection of SSEG (2018)
Available at: https://www.sseg.org.za/wp-content/uploads/2019/03/Grid-
Impact-Studies-P13162_SSEG_Guideline_National_V1.pdf

SSEG municipality Online resource portal that contains information on technical support for
resource portal municipalities to assist them with developing processes for safely
integrating SSEG into municipal grids.
Available at: https://www.sseg.org.za/

CSIR Load Flow Tool The CSIR developed an Excel-based simplified load flow tool with the
main purpose to perform simple load flow on relatively small networks
(including small network equivalents of aggregated, larger networks).
For more information, please contact the CSIR Energy Centre
(https://www.csir.co.za/energy-research-centre).

CSIR Hosting Capacity Tool The CSIR developed a hosting capacity method for South African
municipalities to determine the potential impacts that may occur when
increasing SSEG interconnections to the distribution network. The
application of the tool is in DigSILENT.
For more information, please contact the CSIR Energy Centre
(https://www.csir.co.za/energy-research-centre).

Report for
Doc No: Rev Page 85 of 88
Report for
Doc No: Rev Page 86 of 88
8 References
[1] CIGRE WG C6.24, “Capacity of Distribution Feeders for Hosting Distributed Energy Resources,” 2014.
[2] USAID/Jordan, “Grid Impact Analysis Study Training (GIS).” .
[3] The City of Cape Town, “Solar PV information sheet: SSEG tariff,” 2018. doi: 10.1007/978-1-137-
00472-7_3.
[4] The City of Cape Town, “GUIDELINES FOR EMBEDDED Application process to become an embedded
generator in the City of Cape Town,” 2015.
[5] CSIR, “The CSIR in brief,” 2021. .
[6] J.Calitz;J.Wright, “Statistics of utility-scale power generation in South Africa in 2020.” 2020.
[7] Department of Energy Republic of South Africa, “2019 Energy Sector Report,” 2019. [Online]. Available:
http://dx.doi.org/10.1016/j.powtec.2017.06.008%0Ahttps://doi.org/10.1016/j.jwpe.2017.08.012%0Ahtt
ps://doi.org/10.1016/j.jhazmat.2021.125853%0Ahttps://doi.org/10.1080/10962247.2018.1521349%0A
http://www.energy.gov.za.
[8] South African Local Government Association (SALGA), “Status of Small Scale Embedded Generation
(Sseg) in South African Municipalities 2020,” no. November, pp. 1–24, 2020.
[9] Department of Energy Republic of South Africa, “Integrated Resource Plan,” 2019.
[10] Cliff Dekker Hofmeyer, “ENERGY ALERT The Integrated Resource Plan 2019 : A promising roadmap
for generation capacity in South Africa,” 2019.
[11] NERSA, “Electricity Licences.” https://www.nersa.org.za/electricity-overview/electricity-licences/
(accessed Aug. 03, 2021).
[12] P. Jiang, Y. Van Fan, and J. J. Klemeš, “Impacts of COVID-19 on energy demand and consumption:
Challenges, lessons and emerging opportunities,” Appl. Energy, vol. 285, no. December 2020, 2021,
doi: 10.1016/j.apenergy.2021.116441.
[13] IEA, “Global Energy Review 2020,” 2020. doi: 10.1787/a60abbf2-en.
[14] M. Boulle and A. Dane, “The impacts of Covid-19 on the power sector in sub-Saharan Africa, and the
role of the power sector in socio-economic recovery,” no. July, pp. 1–16, 2020, [Online]. Available:
https://www.kas.de/documents/282730/8327029/Covid_Energy_SSA_publication.pdf/efc74763-8f85-
39c6-53e4-de16cb75f71d?t=1594778811782.
[15] GreenCape, “Small Scale Embedded Generation : Registration vs . application for connection.” pp. 1–
4, 2021.
[16] Presidency of South Afrrica, “Announcement By President Cyril Ramaphosa on Amendment To
Schedule Two of the Electricity Regulation Act,” pp. 1–3, 2021, [Online]. Available:
https://www.groundwork.org.za/Documents/energy/210610_Presidential_Energy_Reforms_Announce
ment.pdf.
[17] Department of Mineral Resources and Energy, “Electricity Regulation Act, 2006 (Act No. 4 of 2006):
Amendment of Notice: Licensing Exemption and Registration Notice,” Government Gazette, Republic
of South Africa, vol. 674. 2021, [Online]. Available: http://www.greengazette.co.za/pages/national-
gazette-37230-of-17-january-2014-vol-583_20140117-GGN-37230-003.
[18] The City of Cape Town, “Requirements for small-scale embedded generation,” 2017.
[19] National Energy Regulator of South Africa (NERSA), “Registration procedure for small-scale embedded

Report for
Doc No: Rev Page 87 of 88
generators (version 1),” no. August, pp. 1–13, 2018, [Online]. Available:
http://www.nersa.org.za/Admin/Document/Editor/file/Electricity/Application Form/Internal Registration
Procedure For Small Scale Embedded Generation.pdf.
[20] SABS, “NRS 097-2-3 : 2014 GRID INTERCONNECTION OF EMBEDDED Part 2 : Small-scale
embedded generation Section 3 : Simplified utility connection criteria for low-voltage connected
generators,” p. 27, 2014.
[21] P. Lilje, “Recommended practice for assessing the connection of small generators based on renewable
energy sources to low-voltage and medium-voltage municipal grids,” 2018.
[22] G. Botha and C. Carter-brown, “Local specifications for small-scale embedded generators,” no. June,
pp. 64–66, 2014.
[23] B. Magoro and T. Khoza, “GRID CONNECTION CODE FOR RENEWABLE POWER PLANTS ( RPPs
) CONNECTED TO THE ELECTRICITY TRANSMISSION SYSTEM ( TS ) OR THE DISTRIBUTION
SYSTEM ( DS ) IN SOUTH AFRICA,” vol. 8, no. July, 2014.
[24] SABS, “NRS-097-2-1: Grid Interconnection of embedded generation section 1: Utility Interface,” 2017,
[Online]. Available: https://scot.eskom.co.za.
[25] Presidency of South Afrrica, “Announcement By President Cyril Ramaphosa on Amendment To
Schedule Two of the Electricity Regulation Act,” pp. 1–3, 2021.
[26] Eskom, NRS 048-2: Electricity Supply - Quality of Supply Part 2 : Voltage characteristics , compatibility
levels , limits and assessment methods. 2017.
[27] Academy of EMC, “EMC Standards.” https://www.academyofemc.com/emc-standards (accessed Jun.
23, 2021).
[28] T. Sathiyanarayanan and M. Sydulu, “Load Flow Analysis for Radial and Mesh Connected Distribution
Systems,” Int. J. Sci. Res., vol. 3, no. 6, pp. 522–528, 2014.
[29] F.Oloo, “GIZ / CSIR WP2 - SSEG Grid Impact Assessments Support for Municipalities : Simplified load
flow assessment tool,” 2021.
[30] A. Mikulec and V. Mikulicic, “Influence of Renewable Energy Sources on Distribution Network
Availability,” Dev. power Eng. Croat., p. 12, 2011.
[31] M. Zain ul Abideen, O. Ellabban, and L. Al-Fagih, “A Review of the Tools and Methods for Distribution
Networks ’ Hosting Capacity Calculation,” pp. 1–25, 2020.
[32] Electric Power Research Institute, “Distribution Feeder Hosting Capacity: What matters when planning
for DER,” 2015.
[33] U. Siegfriedt and C. Brandt, “Solar PV Installation Guidelines.” South African Photovoltaic Indusrty
Association, Johannesburg, p. 77, 2017.
[34] DIgSILENT GmbH, “PowerFactory 2019 - DPL Function Reference,” p. 1269, 2019.
[35] Y. Tan, “Chapter 10 - GPU-Based Random Number Generators,” Y. B. T.-G.-B. P. I. of S. I. A. Tan, Ed.
Morgan Kaufmann, 2016, pp. 147–165.
[36] SABS, “SANS 60950-1 : 2014 IEC 60950-1 : 2013 SOUTH AFRICAN NATIONAL STANDARD
Information technology equipment — Safety Part 1 : General requirements,” p. 288, 2014.
[37] City of Cape Town, “Requirements for small-scale embedded generation,” pp. 1–45, 2017, [Online].
Available: http://resource.capetown.gov.za/documentcentre/Documents/Procedures, guidelines and
regulations/Requiremenst for Samll-Scale Embedded Generation.pdf.

Report for
Doc No: Rev Page 88 of 88
[38] SANS, “SANS 10142-1-2 : 201X DRAFT : SOUTH AFRICAN NATIONAL STANDARD The wiring of
premises Part 1-2 : Specific requirements for embedded generation installations connected to the low
voltage distribution network in South Africa,” 2002.
[39] SABS, “SANS 62040-1 : 2013 SOUTH AFRICAN NATIONAL STANDARD Uninterruptible power
systems ( UPS ) Part 1 : General and safety requirements for UPS,” p. 65, 2023.
[40] SABS, “NRS-097-2-1: Grid Interconnection of embedded generation section 1: Utility Interface,” 2017.

Report for
Doc No: Rev Page 89 of 88

You might also like