giz capacity creditstudiessa_final

85
Capa Final Febru acity Credit of Wind Generation i l Report uary 2011 in South Africa

Upload: afsea

Post on 08-May-2015

136 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Giz capacity creditstudiessa_final

Capacity Credit of Wind Generation in South Africa

Final Report

February 2011

Capacity Credit of Wind Generation in South Africa

Final Report

February 2011

Capacity Credit of Wind Generation in South Africa

Page 2: Giz capacity creditstudiessa_final

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0

Commissioned by:

Consultants:

0 3 4 . 0 0

Deutsche Gesellschaft für Internationale

Zusammenarbeit (GIZ) GmbH

Office Pretoria

Focal Area Energy & Climate

333 Grosvenor Street

Pretoria 0028

South Africa

Contact:

Daniel Werner (daniel.werner@gi

Department of Energy

Contact:

Ompi Aphane ([email protected]

André Otto ([email protected]

Eskom

Contact:

Kevin Leask ([email protected])

DIgSILENT GmbH

Heinrich-Hertz-Str. 9

72810 Gomaringen

Germany

Web:http://www.digsilent.de

Contact:

Markus Pöller ([email protected])

Windlab Developments South Africa Pty Ltd

9b Bell Crescent Close

Westlake 7945

South Africa

Contact:

Francis Jackson (francis.jackson

2

Deutsche Gesellschaft für Internationale

Daniel Werner ([email protected])

energy.gov.za)

energy.gov.za)

Kevin Leask ([email protected])

Markus Pöller ([email protected])

Windlab Developments South Africa Pty Ltd

[email protected])

Page 3: Giz capacity creditstudiessa_final

T a b l e o f C o n t e n t s

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 3

Table of Contents

1 Introduction ......................................................................................................................................... 5

2 Part 1 – Capacity Credit Studies .......................................................................................................... 7

2.1 Approach and Methodology ..................................................................................................................... 7

2.1.1 Generation Adequacy ......................................................................................................................... 7

2.1.2 Consideration of Correlation Effects - Methodology of the Presented Studies ........................................... 8

2.1.3 Assignment of Capacity Credit to Individual Regions or Wind Farms ..................................................... 10

2.1.4 Modelling ........................................................................................................................................ 12

2.1.4.1 Load Model ................................................................................................................................ 12

2.1.4.2 Modelling of Dispatchable Generation ........................................................................................... 14

2.1.4.3 Modelling of Wind Generation ...................................................................................................... 14

2.1.4.3.1 Wind Speed Data ................................................................................................................... 14

2.1.4.3.2 Transformation into Power Time Series Data ............................................................................ 15

2.1.5 Simulation Model ............................................................................................................................. 16

2.2 Scenario Definition ............................................................................................................................... 18

2.2.1 Scenario 1 – Year 2015 with 2000MW of Installed Wind Capacity ........................................................ 18

2.2.2 Scenario 2 – Year 2020-Low Wind Scenario with 4800MW of Installed Wind Capacity ............................ 18

2.2.3 Scenario 3– Year 2020-High Wind Scenario with 10 000MW of Installed Wind Capacity ......................... 19

2.3 Results of Simulation Studies................................................................................................................. 20

2.3.1 Scenario 1 – Year 2015 with 2000MW of Installed Wind Capacity ........................................................ 20

2.3.2 Scenario 2 – Year 2020-Low Wind Scenario with 4800MW of Installed Wind Capacity ............................ 22

2.3.3 Scenario 3– Year 2020-High Wind Scenario with 10 000MW of Installed Wind Capacity ......................... 23

2.3.4 Outlook for up to 25 000MW of Installed Wind Capacity ...................................................................... 24

2.3.5 Wind Farms in the Western and Southern Region ............................................................................... 26

2.3.6 Comparison with Thermal or Hydro Power Plants ............................................................................... 27

3 Part 2: Impact of Wind Generation in South Africa on System Operation ....................................... 28

3.1 Approach and Methodology ................................................................................................................... 28

3.2 Results of Time Series Simulation Studies ............................................................................................... 29

3.2.1 Impact on Main System Performance Characteristics .......................................................................... 29

3.2.2 Local versus Global Wind Generation ................................................................................................. 31

3.2.3 Energy Production and CO2 Emissions ............................................................................................... 32

4 Conclusions and Recommendations .................................................................................................. 33

5 References ......................................................................................................................................... 35

A N N E X E S ...................................................................................................................... 36

Annex 1: Results of Simulation Studies ............................................................................................... 37

Annex 1.1: Scenario 1 – Year 2015 – 2000MW Installed Wind Generation ....................................................... 37

Page 4: Giz capacity creditstudiessa_final

T a b l e o f C o n t e n t s

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 4

Annex 1.2: Scenario 2 – Year 2020 – 4800MW Installed Wind Generation ....................................................... 40

Annex 1.3: Scenario 3 – Year 2020 – 10000MW Installed Wind Generation ..................................................... 43

Annex 2: Seasonal and Hourly Variations of the Wind Generation in South Africa ............................ 46

Annex 2-1: Scenario 1 – Year 2015/2000MW of Installed Wind Capacity ......................................................... 47

Annex 2-2: Scenario 2 – Year 2020-Low/4800MW of Installed Wind Capacity .................................................. 49

Annex 2-3: Scenario 3 – Year 2020-High/10000MW of Installed Wind Capacity ............................................... 51

Annex 3: Results of Time Series Assessment ....................................................................................... 53

Annex 3-1: Worst Case Situations and Duration Curves, Scenario 1 – 2015 ..................................................... 53

Annex 3-2: Worst Case Situations and Duration Curves, Scenario 2 – 2020 ..................................................... 64

Annex 3-3: Worst Case Situations and Duration Curves, Scenario 3 – 2020 ..................................................... 75

Page 5: Giz capacity creditstudiessa_final

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 5

1 Introduction

Renewable energies, such as wind energy, allow electricity production without consuming fossil resources and

without any direct carbon dioxide emissions. Just by producing electrical energy, the use of these sources is

justifiable and represents in many locations an economical alternative to the use of fossil resources such as coal

or oil. Particularly the cost of wind generation has considerably decreased over the last decade and while fossil

fuel prices are increasing it is anticipated that wind generation at high wind sites will represent the most

economical method for generating electrical energy in the near future.

However, in contrast to more conventional power plants based on fossil resources, electricity production from

wind farms cannot be planned because of the variable nature of wind speeds. Therefore, and because there

aren’t any suitable electricity storage technologies available (yet), conventional, dispatchable power plants using

fossil fuels are always required for times during which electricity demand is high and wind generation is low.

This leads to the question about how much conventional generation capacity is still required in a system with high

amount of wind generation. Is it the same amount as without wind generation or is it less? If it was the same

amount, there would still be the benefit of consuming less fossil resources and reducing carbon dioxide emissions

compared to a scenario without wind energy use, but there wouldn’t be any positive influence on the installed

conventional generation capacity. However, in the case that less conventional generation capacity would have to

be installed, this capacity effect could be seen as an additional benefit from wind generation and would make this

technology even more economical. In this case, a “Capacity Credit” could be assigned to wind generation

meaning that a certain percentage of the installed wind generation capacity could be considered as a contribution

to the firm capacity of a system that is required for ensuring a safe and reliable electricity supply.

Even in systems not using renewable energy generation, it is necessary that the installed generation capacity

exceeds the peak demand because also conventional power plants are not permanently available because of

planned outages (for maintenance) and unplanned outages (faults). The excess of installed capacity over the

peak demand is the Reserve Margin and represents an important constraint in the capacity planning process.

According to [6], a reserve margin of 15% has been considered to be appropriate for ensuring a sufficiently

reliable electricity supply of South Africa.

With the addition of wind generation, the classical, empirical approach based on the consideration of a reserve

margin on the basis of typical availability indices of conventional power plants does not work anymore for

assessing the required installed capacity:

• When applying the same reserve margin to the total generation capacity, including wind generation, the

required installed capacity will be highly underestimated because the availability of wind generation is

much below conventional generation based on fossil fuels.

• When applying the same reserve margin to conventional generation capacity only, hence considering

that wind generation has a capacity credit of zero, the required installed capacity will be overestimated

because the installed wind generation will tend to improve the reliability of supply, an effect which

would be ignored by such a simplified approach.

Hence, the question arises how the capacity credit of wind generation can actually be quantified and used in the

capacity planning process for determining the required electricity generation capacity.

This report presents the results of simulation studies for assessing the capacity credit of planned wind farms in

South Africa. Therefore, three scenarios have been defined and agreed between ESKOM, GTZ and DIgSILENT

[1]:

Page 6: Giz capacity creditstudiessa_final

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 6

- Scenario 1: Year 2015: 2 000 MW of installed wind generation capacity

- Scenario 2: Year 2020 – low wind generation scenario: 4 000 MW of installed wind generation capacity

- Scenario 3: Year 2020 – high wind generation scenario: 10 000 MW of installed wind generation capacity

Based on Scenario 3, additional simulations have been carried out looking at up to 25 000 MW of installed wind

generation capacity while keeping the specific distribution of wind generation constant.

The three scenarios are based on realistic assumptions with regard to wind farm sites, the potential installed wind

generation capacity at these sites and the relevant characteristics of the South African power system, such as

load and generation characteristics and planned expansions of thermal and hydro power plants, which have been

provided by ESKOM [3].

A second part of the presented studies is analyzing simulated time series of load and wind generation for the

three above defined scenarios.

This part of the studies is mainly looking at the following aspects:

• Worst case situations with regard to wind generation, load.

• Worst case situations with regard to wind and load variations (ramp-up/ramp-down speeds).

• Impact of wind generation on the cumulative probability curve of the residual load, which has to be

covered by thermal and hydro power plants.

• Impact on residual load variations.

• Energy production and avoided carbon dioxide emissions.

The results of these studies will help determining the impact of wind generation on the required dynamic

performance of thermal and hydro power plants.

Page 7: Giz capacity creditstudiessa_final

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 7

2 Part 1 – Capacity Credit Studies

2.1 Approach and Methodology

2.1.1 Generation Adequacy

Figure 1: Generation Availability Curve

For assessing the capacity credit of wind generation, probabilistic methods for analyzing generation adequacy

have to be applied. The basic concepts can best be explained using the generic example according to Figure 1.

The blue curve in Figure 1 represents the cumulative probability curve for the available generation in a power

system. The x-axis depicts probabilities in %, the y-axis shows the available capacity in MW. The blue curve

shows the minimum available capacity at a specified probability level.

The available generation depends on the following parameters:

• Pinst: The installed capacity of the system

• Pplanned: Planned capacity during the observation period. The planned capacity is equal to the installed

capacity minus planned outages (maintenance).

• Finally, the cumulative probability of the available capacity can be calculated on basis of the unplanned

outages of all power plants in the system.

Using this approach, the demand that can be supplied with sufficient reliability (Psecure) is defined by the

capacity that is at least available at a given confidence level (e.g. 99%).

0 20 40 60 80 100

Pinst

Psecure

Pplanned

Confidence

Reserve Margin

Probability in %

Ca

pa

city

in M

W

Page 8: Giz capacity creditstudiessa_final

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 8

Figure 2: Generator Availability Curve - With Wind

For analyzing the influence of wind generation on the firm capacity the cumulative probability curve of the total

available capacity, including conventional power plants and wind power plants, as depicted in Figure 2 needs to

be calculated. The two curves have the following meaning:

• Blue curve: Cumulative probability curve of conventional generation.

• Green curve: Cumulative probability curve of conventional and wind generation together.

When comparing the green curve and the blue curve, it can be observed that the load level that can be supplied

at the given confidence level has increased. This increase of securely supplied load is also named “Equivalent

Load Carrying Capability” (ELCC) of the system and is a useful index for defining the capacity credit of wind

generation.

2.1.2 Consideration of Correlation Effects - Methodology of the Presented Studies

In countries, in which a very distinct peak load situation can be identified, an approach purely based on the peak

load situation might be sufficiently accurate because it can be assumed that only the peak load level contributes

considerably to the average Loss of Load Probability (LOLP). In this case, it is sufficient to calculate the

generation availability during the peak load time (season and hour of day) and to work out the capacity credit of

wind generation purely on the basis of this simplified analysis.

However, in systems, in which the yearly load profile is relatively flat, not only the peak load situation might lead

to loss of load situations but also other load situation may considerably contribute to the average LOLP. A simple

analysis of generation availability at peak load is then inappropriate.

0 20 40 60 80 100

Pinst

Psecure

Pwinst

ELCC

Confidence

Ca

pa

city

in M

W

Probability in %

Page 9: Giz capacity creditstudiessa_final

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 9

For this reason, the approach on which these studies are based on considers generation availability and the

seasonal variation of daily peak loads together and calculates a cumulative probability curve of the available

Reserve, whereas Reserve is defined by the excess of generation capacity over the daily peak load.

The relevant reliability index, on which the capacity credit definition of the presented studies is based on, is the

average loss of load probability (LOLP) for daily peak loads. The average LOLP during peak load hours can be

translated into a figure indicating the expected number of days per year during which the peak load cannot be

covered by the available generation capacity.

The advantages of this approach are the following:

• It provides a realistic measure for system adequacy improvements due to wind generation.

• Considers correlation between load and planned outages.

• Considers correlation between wind speeds and seasonal load variations.

• Considers correlation between wind speeds and daily load variations.

The LOLP at daily peak loads is expressed in % but can also be expressed in terms of an average number of days

per year during which a loss of load situation has to be expected. For example:

• An average LOLP at daily peak load of 1% would be equivalent to an average of 0,01x365=3,65 days

per year during which the peak load cannot be covered by the available generation.

• An average LOLP at daily peak load of 0,1% is equivalent to an average of 0,365 days per year, hence

around 1 day in 2,7 years at which the daily peak load cannot be covered.

The use of the average LOLP at daily peak load requires an assessment that considers the yearly load variation

and the available capacity simultaneously. Instead of a cumulative probability curve of the available generation

capacity, as presented in the previous section, the cumulative probability of the Reserve is calculated. A negative

Reserve indicates a loss of load situation. Consequently, the probability with which the Reserve is lower than zero

represents the average probability with which the daily peak load cannot be covered by the available generation

capacity.

In this study, Capacity Credit is defined in terms of Equivalent Firm Capacity, which represents a firm capacity

(capacity with 100% availability) that has the same influence on the reliability of supply of the system as the

actually installed wind generation. Because the addition of a perfectly firm capacity would increase the Reserve

by exactly the amount of the added firm capacity, the increase of Reserve at a given confidence level is a suitable

measure for assigning an Equivalent Firm Capacity (EFC) to wind generation.

Page 10: Giz capacity creditstudiessa_final

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 1 0

Figure 3: Cumulative Probability of the Available Reserve

Figure 3 shows the cumulative probability of the “Reserve” for a generic system.

According to Figure 3, the average LOLP without wind generation (blue curve) is equal to around 2%, which is

equivalent to around 7,3 loss of load events per year. The LOLP of the system with wind generation (green curve)

is equal to 99,9%, which corresponds to 0,365 loss of load events per year or 1 loss of load event within 2,7

years on average.

In a real system, there is always some operational reserve required for being able to compensate frequency

deviations etc. Therefore, the equivalent firm capacity has not been defined at the zero crossing of the probability

curve of the Reserve in the presented studies, but instead, the increase of available Reserve at a given

confidence level of 99,9% has been taken.

Besides this, only the winter season has been considered, which represents the South African peak load season.

Because the observed time frame corresponds only to half a year, a confidence level of 99,9%, as used in these

studies, corresponds to less than 1 loss of load event within 5 years.

2.1.3 Assignment of Capacity Credit to Individual Regions or Wind Farms

The capacity credit of wind generation is a non-linear function that depends on

• Average generation of wind farms or capacity factor.

• Wind penetration level (ration between installed wind capacity and installed conventional generation

capacity).

• Planned and unplanned outage rates of conventional power plants.

• Load variations.

0 20 40 60 80 100

EFC

100%-LOLP

0

Probability in %

Re

serv

ein

MW

Page 11: Giz capacity creditstudiessa_final

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 1 1

Because of the nonlinear relationship, it is not possible to calculate the capacity credit of individual wind farms or

wind farms in different regions directly. Capacity credit calculations must always consider the whole system.

However, with the following assumptions it is possible to assign portions of the total equivalent firm capacity to

individual wind farms or regions:

It can be shown that the Equivalent Firm Capacity (EFC) of wind generation can very well be approximated by a

formula that decomposes nonlinear and linear parts:

100

)(

×=

⋅=

r

fl

PEFCCC

PavcpCREFC

with:

• EFC: Equivalent Firm Capacity (in MW)

• CR: Capacity Reduction factor

• cp: Wind penetration level (ratio of installed wind capacity and installed conventional generation

capacity)

• Pavfl: Average production during full load period (season and hour per day)

• CC: Capacity Credit (in %)

• Pr: Rated power of totally installed wind generation (in MW)

In this approximate formula, all non-linear parts are lumped into the “Capacity Reduction Factor” CR, which is a

system-wide parameter that mainly depends on the wind penetration level and which is highly independent from

the wind conditions at individual sites or regions.

The dependence of EFC on site-specific parameters, such as the average production during full load hours, is a

linear relationship in this formula and therefore superposition can be applied:

∑∑ =⋅=

i

i

iifl EFCPavcpCREFC )(

The index i represents the individual wind farms or regions of a system. The equivalent firm capacity of each

wind farm or region i can be calculated using:

ifli PavcpCREFC ⋅= )(

A verification of this approach is depicted in Figure 15 on page 27 showing the factor CR in function of the wind

penetration level for the two different study years 2015 and 2020. Based the calculated EFC for each study year

and the average production during full load hours (Pavfl), which can directly be obtained by analyzing the

available wind speed time series data, CR can be calculated using CR(cp)=EFC/ Pavfl.

As it can be seen in Figure 15, the two curves representing the two study years are lying almost on top of each

other even if wind conditions, number and size of the modelled wind farms are highly different. This confirms the

assumption that CR is mainly independent from site specific data and only depends on system characteristics.

Besides regional allocation of capacity credit of wind generation, the simplified formula allows an easy prediction

of capacity credit and equivalent firm capacities for variations with regard to the wind scenarios (see section 2.2).

Page 12: Giz capacity creditstudiessa_final

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0

In the case that the actual wind farm development will considerably deviate from the assumptions on which the

scenarios of the presented studies were based, it will still be possible to calculate equivalent firm capacity and

capacity credit of wind generation by using the approximate formula and the capacity credit reduction factor CR

in function of the wind penetration level according to

production of the considered wind farms.

2.1.4 Modelling

2.1.4.1 Load Model

The load data used for the presented studies correspond to the “Moderate Load Growth Scenario” of the IRP2010

[6]. Further, the load model considers the Net Demand forecast that has been made for 2015 and 2020

considering the regular load forecast and the effect of Demand Side Managemen

implemented until 2015 and 2020.

Figure 4: Forecast of Daily Peak Load

0 3 4 . 0 0

In the case that the actual wind farm development will considerably deviate from the assumptions on which the

narios of the presented studies were based, it will still be possible to calculate equivalent firm capacity and

capacity credit of wind generation by using the approximate formula and the capacity credit reduction factor CR

ion level according to Figure 15 and to multiply the CR with the average

production of the considered wind farms.

ented studies correspond to the “Moderate Load Growth Scenario” of the IRP2010

[6]. Further, the load model considers the Net Demand forecast that has been made for 2015 and 2020

considering the regular load forecast and the effect of Demand Side Management measures that will be

implemented until 2015 and 2020.

: Forecast of Daily Peak Load (Net Demand) for 2015 and 2020

1 2

In the case that the actual wind farm development will considerably deviate from the assumptions on which the

narios of the presented studies were based, it will still be possible to calculate equivalent firm capacity and

capacity credit of wind generation by using the approximate formula and the capacity credit reduction factor CR

and to multiply the CR with the average

ented studies correspond to the “Moderate Load Growth Scenario” of the IRP2010

[6]. Further, the load model considers the Net Demand forecast that has been made for 2015 and 2020

t measures that will be

Page 13: Giz capacity creditstudiessa_final

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 1 3

Figure 5: Typical hourly load variations in winter

Figure 6: Typical hourly load variations in summer

More precisely, the load model is based on the following load data according to [3]:

• Forecast of daily peak load for 2015 and 2020 (see Figure 4)

• Typical relative hourly load variations for a typical week in winter (see Figure 5) and a typical week in

summer (see Figure 6).

Based on the typical hourly load variations (see Figure 5 and Figure 6), the following information has been

extracted and used by the studies:

• In winter, full load hours are between 18:00h and 21:00h.

• In summer, the daily load is relatively flat and full load hours can occur between 08:00h and 21:00h.

0

0,2

0,4

0,6

0,8

1

1,2

1 10 19 28 37 46 55 64 73 82 91 100 109 118 127 136 145 154 163

Hour per week

rela

tive

lo

ad

0

0,2

0,4

0,6

0,8

1

1,2

1 10 19 28 37 46 55 64 73 82 91 100 109 118 127 136 145 154 163

Hour per week

rela

tive load

Page 14: Giz capacity creditstudiessa_final

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 1 4

2.1.4.2 Modelling of Dispatchable Generation

For modelling the availability of dispatchable power plants, ESKOM provided planned and unplanned outage rates

of all South African power stations [3]:

Unplanned outages are modelled by two-state Markov-models considering their unplanned outage rates as

provided by ESKOM.

For considering planned outages (maintenance) the monthly system-wide planned outage rate has been provided

by ESKOM [3]. Based on this information, a number of generators have been “put on maintenance” in each

month. The total number of generators on outage in each month considers the maximum limits that were

provided by ESKOM.

2.1.4.3 Modelling of Wind Generation

2.1.4.3.1 Wind Speed Data

Windlab Systems has carried out studies [4] for producing time series data of hourly average wind speeds at 80m

height at all potential wind farm sites, which have been defined by ESKOM [2],

The Windlab methodology is based on an atmospheric modelling approach whose input are synoptic weather data

gathered by the world meteorological organisation (WMO).

Based on these data a regional scale model is initially applied allowing modelling wind speeds with a 6 hours/1

degree resolution.

This regional model accounts for forces associated with stratification and inertial forces, friction with the earth’s

surface, accelerations and steering over and around large scale topographical features (such as mountains) and

thermal circulations such as sea breezes.

The fine-scale model then uses the regional-scale model as a boundary condition to create a yet finer grid,

typically of 100 meter resolution. This model takes into account friction from various types of land cover (forests,

crops etc) and acceleration over and around smaller-scale topography.

The fine-scale model is then validated using Automatic Weather Stations (AWS) and Meteorological mast data,

where available. This validation is applied to the model to reflect the long-term average.

Time-series of wind speed with a time resolution of 1 hour and direction at the points of interest are extracted

from the model. These allow the generation of statistical tools such as wind roses, probability distributions and

monthly averages. These time-series are point statistics and unless great attention has been paid to the selection

of these locations they cannot be used to represent the absolute wind resource of a region such as a wind farm

or collection of wind farms, for instance for assessment of individual project feasibility.

These time series data of wind speeds have been calculated for each potential wind farm site as defined by

ESKOM for the different scenarios and were used as input into the capacity credit studies.

Page 15: Giz capacity creditstudiessa_final

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 1 5

2.1.4.3.2 Transformation into Power Time Series Data

Figure 7: Generic power curve used for Capacity Credit Studies

For transforming wind speed data into power data, a generic power curve was considered (see Figure 7).

Losses or technical unavailability have not been considered explicitly because these effects are covered by the

general uncertainty margin that has to be expected by the study results.

For accurately considering correlation between wind generation and load, time series of wind speeds have been

used and converted into a power time series (see e.g. Figure 8) at each site using the power curve according to

Figure 7. The daily and seasonal correlation between wind speeds and load is therefore automatically considered

by the corresponding time series data.

The diurnal correlation between wind speed and load is considered by only using wind speed data during full load

hours (see also Figure 8). As described in section 2.1.3, these full load hours are different for summer and winter.

For this reason, two capacity credit calculations have been carried out for each scenario, one for the summer

season and one for the winter season.

0

0,2

0,4

0,6

0,8

1

1,2

0 5 10 15 20 25 30

wind speed in m/s

power output in p.u.

Page 16: Giz capacity creditstudiessa_final

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 1 6

Figure 8: Example of Wind Generation Time Series as used for studies (red line)

The following wind speed data were considered for each season:

• Winter season: Wind speed time series of the months April to September considering only wind speeds

during 18:00h and 21:00h every day.

• Summer season: Wind speed time series from October to March considering only wind speeds during

08:00h and 21:00h every day.

Because the winter load is considerably higher than the summer season it can be assumed that events in summer

have only a very low contribution to the LOLP. Therefore, only the results obtained for the winter season have

finally been used for assessing capacity credit of wind generation.

2.1.5 Simulation Model

Capacity credit studies for the South African case have been carried out by the use of 3 main scenarios,

representing different study years and different assumptions with regard to the installed wind generation

capacity.

The DIgSILENT PowerFactory model integrates all conventional generators, all wind farms as defined for each

scenario and the system load represented by a daily peak load characteristic.

4440,04426,04412,04398,04384,04370,0

1700,00

1400,00

1100,00

800,00

500,00

200,00

44000,00

40000,00

36000,00

32000,00

28000,00

24000,00

x-Axis: TimeSeries: Hour of year in h

TimeSeries: Total Wind Generation in MW

TimeSeries: Total Load in MW

DIg

SIL

EN

T

Page 17: Giz capacity creditstudiessa_final

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 1 7

For calculating the capacity credit of wind generation, a Monte Carlo analysis approach has been used,

considering:

• Daily peak load characteristic.

• Planned and unplanned outages of conventional generators.

• Correlation of wind speeds at different sites.

• Daily, weekly and monthly correlation between wind speeds and the daily peak load.

• Correlation between wind speeds and daily full load hours.

The Monte Carlo Simulation model calculates the following key quantities:

• Cumulative probability curve of the available conventional generation capacity

• Cumulative probability curve of daily peak-loads (peak-load duration curve)

• Cumulative probability curve of the residual load (load minus wind generation)

• Cumulative probability curve of the “Reserve” (generation – load)

The average loss of load probability at daily peak load can be obtained from the cumulative probability of the

Reserve: Every case, in which the Reserve is less than zero indicates that the available generation capacity is not

able to cover the load at this point. An evaluation of the cumulative probability curve of the Reserve allows

quantifying the probability of such a loss of load situation.

Page 18: Giz capacity creditstudiessa_final

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 1 8

2.2 Scenario Definition

For analyzing the contribution of wind generation in South Africa to the firm capacity of the system, three main

scenarios for different study years and different assumptions with regard to the prospected installed wind

generation capacity have been defined:

• Scenario 1: Year 2015 with 2000MW of wind generation

• Scenario 2: Year 2020 with 4800MW of wind generation (low wind scenario)

• Scenario 3: Year 2020 with 10000MW of wind generation (high wind scenario)

2.2.1 Scenario 1 – Year 2015 with 2000MW of Installed Wind Capacity

Scenario 1 applies to the year 2015. The assumed installed wind generation capacity is equal to 2000MW.

The main parameters of Scenario 1 are the following:

• Total installed conventional (thermal, hydro, pump storage and nuclear) capacity: 52 537MW

• Installed wind generation capacity – total: 2 000MW

• Installed wind generation capacity in Western Region: 1 550MW

• Installed wind generation capacity in Southern Region: 450MW

• Peak load (Peak Net Demand): 40 582MW

• Wind penetration level based on installed conventional capacity: 3,8%

• Wind penetration level based on peak load 4,9%

2.2.2 Scenario 2 – Year 2020-Low Wind Scenario with 4800MW of Installed Wind Capacity

Scenario 2 applies to the year 2020 and makes rather pessimistic assumptions for the amount of wind generation

that will be installed in South Africa until then. The main parameters of Scenario 2 are:

• Total installed conventional (thermal, hydro, pump storage and nuclear) capacity: 59 753MW

• Installed wind generation capacity – total: 4 800MW

• Installed wind generation capacity in Western Region: 3 200MW

• Installed wind generation capacity in Southern Region: 1 600MW

• Peak load (Peak Net Demand): 48 316MW

• Wind penetration level based on installed conventional capacity: 8,0%

• Wind penetration level based on peak load 9,9%

Page 19: Giz capacity creditstudiessa_final

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 1 9

2.2.3 Scenario 3– Year 2020-High Wind Scenario with 10 000MW of Installed Wind Capacity

Scenario 3 is identical to Scenario 2 with regard to the conventional generation and the load data but making

more optimistic assumptions with regard to the wind generation capacity that will be installed until 2020. The key

parameters of scenario 3 are:

• Total installed conventional (thermal, hydro, pump storage and nuclear) capacity: 59 753MW

• Installed wind generation capacity – total: 10 000MW

• Installed wind generation capacity in Western Region: 7 800MW

• Installed wind generation capacity in Southern Region: 2 200MW

• Peak load (Peak Net Demand): 48 316MW

• Wind penetration level based on installed conventional capacity: 16,7%

• Wind penetration level based on peak load 20,7%

The wind penetration level of the three scenarios varies between around 5% and 20% (based on peak load),

which can be considered to be moderate, even in Scenario 3.

Page 20: Giz capacity creditstudiessa_final

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 2 0

2.3 Results of Simulation Studies

For analyzing the capacity credit of planned wind farms in South Africa, Monte-Carlo simulations using the power

system analysis software DIgSILENT PowerFactory have been carried out applying the methods described in

section 2.1.

The increase of the available Reserve at a confidence level of 99,9% has been defined to be the Equivalent Firm

Capacity of wind generation in South Africa (see section 2.1). The results of the capacity credit assessment for

the three scenarios are presented in the following sections.

2.3.1 Scenario 1 – Year 2015 with 2000MW of Installed Wind Capacity

Figure 9: Available Reserve with and without wind generation for Scenario 1 (range > 99%)

Figure 9 shows the cumulative probability curve of the available Reserve in the South African System with wind

generation (red) and without (blue) wind generation for the winter (high load) season.

At the given confidence level of 99,9%, the difference between the two curves is equal to 536MW, which

corresponds to a capacity credit of 26,8%.

Additional results of capacity credit simulations for the winter 2015 case are depicted in Annex 1.1.

The key results for Scenario 1 (2015) are the following:

100,0099,8099,6099,4099,2099,00 [%]

8000,00

6000,00

4000,00

2000,00

0,00

-2000,00

Summary Grid: Total Reserve Generation (Unconstrained) in MW

Summary Grid: Reserve Dispatchable Generation (Unconstrained) in MW

X = 99,900 %

2251.571 MW

2786.828 MW

0.000 MW

Y =

0

,00

0 M

W

99.998 %

DIg

SIL

EN

T

Page 21: Giz capacity creditstudiessa_final

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 2 1

• Available Reserve at 99,9% confidence without wind: R=2 252 MW

• Available Reserve at 99,9% confidence with wind: R=2 787 MW

• Equivalent Firm Capacity: EFC= 535 MW

• Capacity Credit: CC= 26,8%

• Average Wind Generation (year): Pav= 543 MW

(Av. Capacity Factor: 27,17%)

• Average Wind Generation during full load hours: Pav_full= 605 MW (30,25% of Pwinst)

Because the wind penetration level is still relatively low in Scenario 1, the capacity credit of wind generation is

very close to the average capacity factor of wind generation in Scenario 1.

Because of the low wind penetration level (around 5% based on peak load), the Capacity Reduction Factor CR is

very close to 0,9 (see also Figure 15). On the other hand, the average wind generation during full load hours is

considerably higher than the yearly average wind generation. These two aspects – Capacity Reduction Factor and

positive correlation between wind generation and load – almost compensate each other leading to a capacity

credit, which is almost equal to the capacity factor.

Page 22: Giz capacity creditstudiessa_final

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 2 2

2.3.2 Scenario 2 – Year 2020-Low Wind Scenario with 4800MW of Installed Wind Capacity

Figure 10: Available Reserve with and without wind generation for Scenario 2 (range > 99%)

The cumulative probability curve of the available Reserve with and without wind generation of Scenario 2 is

depicted in Figure 10. At the given confidence level of 99,9% the difference between the Reserve with wind

generation and without wind generation is here equal to 1218 MW, which is equal to 25,4% of the installed wind

generation capacity of this scenario.

Additional results of capacity credit simulations for the winter 2020/low wind generation case are depicted in

Annex 1.2.

The key results for Scenario 2 (2020-low wind generation) are the following:

• Available Reserve at 99,9% confidence without wind: R=1 459 MW

• Available Reserve at 99,9% confidence with wind: R=2 678 MW

• Equivalent Firm Capacity: EFC=1 218 MW

• Capacity Credit: CC= 25,4%

100,0099,8099,6099,4099,2099,00 [%]

1.20E+4

9.00E+3

6.00E+3

3.00E+3

0.00E+0

-3.00E+3

Summary Grid: Total Reserve Generation (Unconstrained) in MW

Summary Grid: Reserve Dispatchable Generation (Unconstrained) in MW

X = 99,900 %

1459.123 MW

2677.937 MW

0.000 MWY = 0,000 MW99.991 %99.999 %

DIg

SIL

EN

T

Page 23: Giz capacity creditstudiessa_final

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 2 3

• Average Wind Generation (year): Pav=1 468 MW

(Av. Capacity Factor: 30,6%)

• Average Wind Generation during full load hours: Pav_full=1 652 MW (34,4% of Pwinst)

Comparing the capacity credit figures of Scenario 2 with Scenario 1 it can be observed that capacity credit for

Scenario 2 is only slightly below the Scenario 1 capacity credit, even if the wind penetration level increases

considerably (CP around 10% based on peak load), which would normally lead to a reduction of the capacity

credit (see Figure 15).

The reason for this effect is that the average wind speeds of the wind farm sites considered for Scenario 2 are

higher than in Scenario 1 (Av. Capacity Factor of 30,6% compared to 27,2%/Scenario 1). This means that the

sites with the best wind conditions will not necessarily be developed first, but other aspects, such as the

surrounding infrastructure (e.g. the next grid access point) might play an important role in the wind farm

development process too. Additionally, wind farm developers will first develop smaller projects requiring lower

investments and less grid expansion so that some of the very large wind farm projects proposed in high wind

areas in South Africa will be expected to be developed at later stages.

2.3.3 Scenario 3– Year 2020-High Wind Scenario with 10 000MW of Installed Wind Capacity

Figure 11: Available Reserve with and without wind generation for Scenario 3 (range > 99%)

100,0099,8099,6099,4099,2099,00 [%]

5976,0

4550,6

3125,2

1699,9

274,54

-1150,8

Summary Grid: Total Reserve Generation (Unconstrained) in MW

Summary Grid: Reserve Dispatchable Generation (Unconstrained) in MW

X = 99,900 %

1465.463 MW

3768.771 MW

-0.000 MW

Y =

-0,0

00 M

W

99.993 %

DIg

SIL

EN

T

Page 24: Giz capacity creditstudiessa_final

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0

Figure 11 shows the cumulative probability curve of the Available Reserve with and without wind generation for

Scenario 3. The Equivalent Firm Capacity of this scenario is equal to 2 256MW, which is eq

installed wind generation capacity.

Additional results of capacity credit simulations for the winter 2020/high wind generation case are depicted in

Annex 1.3.

The key results for Scenario 3 (2020

• Available Reserve at 99,9% confidence without wind:

• Available Reserve at 99,9% confidence with wind:

• Equivalent Firm Capacity:

• Capacity Credit:

• Average Wind Generation (year

• Average Wind Generation during full load hours:

The average capacity factor of wind farms in this scenario further increase compared to Scenario 2, while t

capacity credit decreases considerably. This is because of the considerably increased wind penetration level

(CP=20,7% based on peak load)

Capacity Reduction Factor CR is down to 0

compared to Scenario 1 and Scenario 2

2.3.4 Outlook for up to 25 000MW of

Figure 12: EFC in function of the installed wind capacity

0 3 4 . 0 0

shows the cumulative probability curve of the Available Reserve with and without wind generation for

The Equivalent Firm Capacity of this scenario is equal to 2 256MW, which is eq

installed wind generation capacity.

results of capacity credit simulations for the winter 2020/high wind generation case are depicted in

The key results for Scenario 3 (2020-high wind generation) can be summarized as follows

Available Reserve at 99,9% confidence without wind: R=1 456 MW

Available Reserve at 99,9% confidence with wind: R=3 716 MW

Equivalent Firm Capacity: EFC=2 256 MW

CC= 22,6%

Average Wind Generation (year): Pav=3 200 MW

(Av. Capacity Factor: 32%)

Average Wind Generation during full load hours: Pav_full=3 577 MW (35,8% of Pwinst)

The average capacity factor of wind farms in this scenario further increase compared to Scenario 2, while t

capacity credit decreases considerably. This is because of the considerably increased wind penetration level

(CP=20,7% based on peak load), which leads to a degradation of capacity credit of wind generation

Capacity Reduction Factor CR is down to 0,63 in this scenario, which explains the degradation of Capacity Credit

compared to Scenario 1 and Scenario 2 (see also Figure 15).

Outlook for up to 25 000MW of Installed Wind Capacity

: EFC in function of the installed wind capacity

2 4

shows the cumulative probability curve of the Available Reserve with and without wind generation for

The Equivalent Firm Capacity of this scenario is equal to 2 256MW, which is equal to 22,6% of the

results of capacity credit simulations for the winter 2020/high wind generation case are depicted in

as follows:

(Av. Capacity Factor: 32%)

Pav_full=3 577 MW (35,8% of Pwinst)

The average capacity factor of wind farms in this scenario further increase compared to Scenario 2, while the

capacity credit decreases considerably. This is because of the considerably increased wind penetration level

, which leads to a degradation of capacity credit of wind generation. The

,63 in this scenario, which explains the degradation of Capacity Credit

Page 25: Giz capacity creditstudiessa_final

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0

Figure 13: Capacity Credit in function of the installed wind capacity

For estimating the dependence of capacity credit on

have been carried out looking at increased wind generation capacities

wind generation constant.

In practical terms this means that

farm is scaled up while keeping all other parameters unchanged.

The results of the sensitivity studies are depicted in

increasing wind penetration levels, the capacity credit of wind generation is reducing.

But even at a level of 25 000MW of installed wind generation capacity

level of around CP=50% (based on the 2020 peak load)

equivalent firm capacity of the South African system is still considerable

Because the Capacity Reduction Factor CR mainly depends on the wind penetration level, it is possible that the

actual Capacity Reduction Factor CR will actually be higher when the installed wind generation capacity will reach

a level as high as 25 000MW because t

credit for a totally installed wind generation capacity of 25 000MW can be considered to be a conservative

estimate.

0 3 4 . 0 0

: Capacity Credit in function of the installed wind capacity

For estimating the dependence of capacity credit on further increasing wind penetration level

at increased wind generation capacities while keeping the specific distribution of

practical terms this means that for the purpose of these sensitivity studies the rating of each individual wind

farm is scaled up while keeping all other parameters unchanged.

studies are depicted in Figure 12 and Figure 13. The results show that with

increasing wind penetration levels, the capacity credit of wind generation is reducing.

000MW of installed wind generation capacity, which corresponds to a wind penetration

level of around CP=50% (based on the 2020 peak load) , the contribution of additional wind generation to the

equivalent firm capacity of the South African system is still considerable and amounts to around 17,6%.

the Capacity Reduction Factor CR mainly depends on the wind penetration level, it is possible that the

actual Capacity Reduction Factor CR will actually be higher when the installed wind generation capacity will reach

a level as high as 25 000MW because the load will grow beyond the 2020 level. Hence, the estimated capacity

credit for a totally installed wind generation capacity of 25 000MW can be considered to be a conservative

2 5

wind penetration levels, additional studies

while keeping the specific distribution of

the rating of each individual wind

. The results show that with

h corresponds to a wind penetration

, the contribution of additional wind generation to the

and amounts to around 17,6%.

the Capacity Reduction Factor CR mainly depends on the wind penetration level, it is possible that the

actual Capacity Reduction Factor CR will actually be higher when the installed wind generation capacity will reach

Hence, the estimated capacity

credit for a totally installed wind generation capacity of 25 000MW can be considered to be a conservative

Page 26: Giz capacity creditstudiessa_final

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0

2.3.5 Wind Farms in the

Figure 14: Capacity Credit of Wind Farms in South Africa (Western and Southern Region)

For assigning a capacity credit to the wind farms of the individual regions, it can be assumed that capacity credit

is approximately proportional to the average win

day). The factor of proportionality is

load probability and the wind penetration level

The results of capacity credit calculations for the different regions

described in section 2.1.3, are depicted in

capacity credit of wind farms in the Southern Region is considerably higher than the average capacity credit of

wind farms in the Western Region, whereas in 2020 (Scenario 2 and 3), the average capacity credit of wind farms

in both regions is almost equal.

This means that the wind farm sites in the Southern Region that have been selected for Scenario 1 have better

wind conditions than the wind farm sites in the Western Region. In the longer term (Scenario 2020) the wind

conditions of wind farms in both regions are very simila

0 3 4 . 0 0

Wind Farms in the Western and Southern Region

Capacity Credit of Wind Farms in South Africa (Western and Southern Region)

For assigning a capacity credit to the wind farms of the individual regions, it can be assumed that capacity credit

is approximately proportional to the average wind generation during the peak load period (season and time of the

day). The factor of proportionality is predominantly a system-wide parameter and depends mainly on the

wind penetration level (see also section 2.1.3).

The results of capacity credit calculations for the different regions that have been calculated using the formulas

are depicted in Figure 14. This figure shows that in Scenario 1/2015, the average

wind farms in the Southern Region is considerably higher than the average capacity credit of

wind farms in the Western Region, whereas in 2020 (Scenario 2 and 3), the average capacity credit of wind farms

e wind farm sites in the Southern Region that have been selected for Scenario 1 have better

wind conditions than the wind farm sites in the Western Region. In the longer term (Scenario 2020) the wind

conditions of wind farms in both regions are very similar.

2 6

Capacity Credit of Wind Farms in South Africa (Western and Southern Region)

For assigning a capacity credit to the wind farms of the individual regions, it can be assumed that capacity credit

d generation during the peak load period (season and time of the

wide parameter and depends mainly on the loss of

that have been calculated using the formulas

in Scenario 1/2015, the average

wind farms in the Southern Region is considerably higher than the average capacity credit of

wind farms in the Western Region, whereas in 2020 (Scenario 2 and 3), the average capacity credit of wind farms

e wind farm sites in the Southern Region that have been selected for Scenario 1 have better

wind conditions than the wind farm sites in the Western Region. In the longer term (Scenario 2020) the wind

Page 27: Giz capacity creditstudiessa_final

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0

Figure 15: Capacity Reduction Factor in function of installed wind generation

The “Capacity Reduction Factor”

load hours is depicted in Figure 15

depends mainly on the following parameters:

• System characteristics (generation/load), especi

• Wind penetration level.

and is (almost) independent from spec

curves for 2015 and 2020 are lying almost on top of each other.

2.3.6 Comparison with Thermal or Hydro Power Plants

The results of the presented capacity credit studies are expressed

means that the contribution of wind generation to the reliability of the South African power system is compared

with a perfect power plant with an availability of 100%

However, actual coal fired power stations

planned and unplanned outages). This means that when comparing wind generation in South Africa with actual

coal fired power stations, the Equivalent Capacity (EC) is between 11% and 25%

Equivalent Firm Capacity. Considering this aspect the following capacity credit figures can be assigned to

generation in South Africa:

• Scenario 1: EFC= 536MW

• Scenario 2: EFC=1218MW

• Scenario 3: EFC=2256MW

These figures mean that the installed capacity of a wind farm must be around 3 to 4 times higher than the

installed capacity of a coal fired power pla

0 3 4 . 0 0

: Capacity Reduction Factor in function of installed wind generation

CR, which is the ratio of capacity credit and average wind generation during full

15 for the winter seasons 2015 and 2020. The “Capacity Reduction Factor” CR

depends mainly on the following parameters:

System characteristics (generation/load), especially the LOLP without wind generation.

is (almost) independent from specific wind site characteristics. This is confirmed by

curves for 2015 and 2020 are lying almost on top of each other.

Comparison with Thermal or Hydro Power Plants

The results of the presented capacity credit studies are expressed in terms of “Equivalent Firm Capacity”. This

means that the contribution of wind generation to the reliability of the South African power system is compared

with a perfect power plant with an availability of 100%.

However, actual coal fired power stations in South Africa have an availability between 80% and 90% (considering

planned and unplanned outages). This means that when comparing wind generation in South Africa with actual

coal fired power stations, the Equivalent Capacity (EC) is between 11% and 25% higher than the reported

Equivalent Firm Capacity. Considering this aspect the following capacity credit figures can be assigned to

MW 596MW < EC < 670MW 29,8% < CC <

MW 1253MW< EC < 1523MW 26,1% < CC < 3

MW 2507MW < EC < 2820MW 25,1% < CC <

These figures mean that the installed capacity of a wind farm must be around 3 to 4 times higher than the

installed capacity of a coal fired power plant in order to have the same effect on generation adequacy.

2 7

: Capacity Reduction Factor in function of installed wind generation

, which is the ratio of capacity credit and average wind generation during full

for the winter seasons 2015 and 2020. The “Capacity Reduction Factor” CR

ally the LOLP without wind generation.

This is confirmed by Figure 15, in which the

in terms of “Equivalent Firm Capacity”. This

means that the contribution of wind generation to the reliability of the South African power system is compared

in South Africa have an availability between 80% and 90% (considering

planned and unplanned outages). This means that when comparing wind generation in South Africa with actual

higher than the reported

Equivalent Firm Capacity. Considering this aspect the following capacity credit figures can be assigned to wind

% < CC < 33,5%

% < CC < 31,7%

% < CC < 28,2%

These figures mean that the installed capacity of a wind farm must be around 3 to 4 times higher than the

generation adequacy.

Page 28: Giz capacity creditstudiessa_final

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 2 8

3 Part 2: Impact of Wind Generation in South Africa on System Operation

3.1 Approach and Methodology

This second part of the presented studies assesses the operational impact of wind generation in South Africa.

The approach of these studies is based on a time series analysis of load and wind generation for the three

scenarios introduced in section 2.2, which are:

• Scenario 1: Year 2015, 2000MW of installed wind generation capacity

• Scenario 2: Year 2020, 4800MW of installed wind generation capacity

• Scenario 3: Year 2020, 10 000MW of installed wind generation capacity

Since the studies presented in this part 2 are only based on an analysis of the residual load, without looking in

detail at dynamic performance characteristics of existing South African power plants, the results presented here

can only give an indication about potential issues and cannot make any definite statement with regard to the

impact of wind generation on operational reserve and dynamic performance requirements of thermal and hydro

power plants in South Africa.

Further studies will be required for simulating the operation of the South African power system under the new

conditions.

This report will present time series data of the following worst case situations:

• Worst-case Wind ramp-up and ramp-down conditions.

• Peak load situations.

• Minimum residual load situations.

These time series simulations are not based on actual measurements but on projected hourly load data for 2015

and 2020 (provided by ESKOM) and wind speed time series of all considered wind farm sites (provided by

Windlab Systems, see section 2.1.4.3.1 and [4]).

Based on these time series data, duration curves (cumulative probability curves) are calculated for wind

generation and load as well as for hourly variations of wind generation and load.

Page 29: Giz capacity creditstudiessa_final

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 2 9

3.2 Results of Time Series Simulation Studies

Worst case situations for each of the three scenarios are depicted in Annex 3. The figures in Annex 3 show the

following situations:

• Typical Winter Situation (long term)

• Typical Summer Situation (long term)

• Week with maximum observed wind generation

• Week with minimum observed wind generation

• Week with maximum observed ramp-up situation

• Week with maximum observed ramp-down situation

• Week during which peak load occurs

• Week during which minimum residual load occurs

• Load/residual load and wind generation duration curves

• Duration curves of hourly variations of load, residual load and wind generation

3.2.1 Impact on Main System Performance Characteristics

Table 1: Worst-case observations of time domain simulations

Sc1 Sc2 Sc3

Peak Net Demand [MW] 40582 48 316 48 316

Installed Wind Capacity [MW] 2000 4800 10000

Max. Wind Generation [MW] 1741 4227 8471

Max. hourly load variation (ramp-up) [MW/h] 3897 4539 4539

Max. hourly load variation (ramp-down) [MW/h] -3845 -4536 -4536

Max. hourly variation of wind generation (ramp-up) [MW/h] 502,4 1369 2547

Max. hourly variation of wind generation (ramp-down [MW/h] -356 -1102 -1670

Max. hourly variation of residual load (ramp-up) [MW/h] 3862 4542 4788

Max. hourly variation of residual load (ramp-down) [MW/h] -3847 -4527 -4797

Page 30: Giz capacity creditstudiessa_final

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 3 0

Figure 16: Cumulative probability of hourly variations of load, wind generation and residual load

(Scenario 3)

Worst-case observations of time series studies carried out for the three scenarios are summarized in Table 1. The

cumulative probability curves of hourly variations of load, wind generation and residual load of Scenario 3 are

depicted in Figure 16.

Based on the cumulative probability curves, which are depicted in Figure 16, the following conclusions with

regard to the impact of up to 10 000MW of installed wind capacity until 2020 can be taken:

• Hourly variations of wind generation (ramp-up and ramp-down) are substantially smaller than hourly

load variations in all three investigated scenarios.

• Consequently, hourly variations of the residual load are almost not affected by wind generation

considered for scenario 1, 2 and 3 (see also Figure 16).

Consequently, required ramp-up and ramp-down speeds of thermal and hydro power plants in South Africa don’t

have to be increased because of the added wind generation.

The main impact on the operation of the South African power system will result from the predictability of wind

generation, which will depend on the quality of wind prediction tools that will be installed for supporting the

operation of the system. Based on the studies presented in this report, it is not possible to assess the impact of

wind generation in South Africa on regulation reserve. For this purpose, more detailed models considering the

actual operation of the South African power system and simulated wind prediction will be required.

25,0020,0015,0010,005,000,00

9000,00

6000,00

3000,00

0,00

-3000,00

-6000,00

x-Axis: TimeSeries: Proability (Duration Curves) in %

TimeSeries: Hourly Load Variation (up) in MW/h

TimeSeries: Hourly Load Variation (down) in MW/h

TimeSeries: Hourly Wind Variation (up) in MW/h

TimeSeries: Hourly Wind Variation (down) in MW/h

TimeSeries: Hourly Residual Load Variation (up) in MW/h

TimeSeries: Hourly Residual Load Variation (down) in MW/h

DIg

SIL

EN

T

Page 31: Giz capacity creditstudiessa_final

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 3 1

3.2.2 Local versus Global Wind Generation

Figure 17: Total wind generation of Scenario 3 and wind generation of single wind farms in the

Western Region (blue) and Southern Region (green).

Figure 17 shows the total wind generation for Scenario 3 (2020/high wind generation) in comparison to the

generation of one arbitrary selected in farm in the Western Region (blue curve) and one arbitrary selected wind

farm in the Southern Region (green curve).

The corresponding analysis leads to the following conclusions:

• It is unlikely that the total wind generation in South Africa exceeds a value of 90% of the installed

capacity whereas the output of individual wind farms can reach 100% of their maximum possible output.

• Variations of the total wind generation are substantially smoother that variations of the output of

individual wind farms.

• There is only a weak correlation between wind farms in the Western and the Southern Region.

The weak correlation between wind speeds in the Southern region and the Western region will lead to smoother

overall wind variations and therefore to an improved overall predictability

2900,002800,002700,002600,002500,002400,00 [-]

1,20

0,90

0,60

0,30

0,00

-0,30

TimeSeries: Total Wind Generation in p.u. (base: 10000,00 MW)

TimeSeries_Wind: Total Wind Generation in p.u. (base: 250,00 MW)

TimeSeries_Wind_South: Total Wind Generation in p.u. (base: 200,00 MW)

DIg

SIL

EN

T

Page 32: Giz capacity creditstudiessa_final

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 3 2

3.2.3 Energy Production and CO2 Emissions

With regard to wind energy production, the following results for the average power (Pav), the capacity factor

(CF) and the yearly energy yield (E) can be estimated:

• Scenario 1: Pav= 543MW CF = 27,2% E=4,76 TWh/year

• Scenario 2: Pav=1468MW CF = 30,6% E=12,86 TWh/year

• Scenario 3: Pav=3200MW CF = 32,0% E=28,03 TWh/year

Assuming that wind generation will mainly replace energy produced by coal fired power plants, the specific

avoided CO2 emissions can be estimated to be in a range of around 1000 kg/MWh. For a more accurate

assessment, one would have to look at the actual generator dispatch in South Africa in the presence of wind

generation and take the reduced efficiency of coal fired power stations into account when letting them operate

below rated output.

Based on 1000 kg/MWh CO2 and not considering any side effects on the efficiency of thermal power stations, the

avoided CO2 emissions can be estimated as follows:

• Scenario 1: 4,76 Mio t CO2

• Scenario 2: 12,86 Mio t CO2

• Scenario 3: 28,03 Mio t CO2

Because of the large number of assumptions, which led to these results, their accuracy is very limited and the

results should only be used for estimating possible order of magnitudes.

Page 33: Giz capacity creditstudiessa_final

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 3 3

4 Conclusions and Recommendations

Renewable energies, such as wind energy can represent an economic and ecologic alternative for the generation

of electrical energy. But besides contributing to the electrical energy supply, renewable energies such as wind

energy can also have a valuable contribution to the equivalent firm capacity of a system. This means in other

words, that with the addition of wind farms, the reliability of supply of a system is improved and that it is indeed

possible to replace some conventional power plants by wind farms completely. The percentage of the installed

wind capacity that can be considered to be “firm capacity” is named Capacity Credit and is subject to the analysis

of the presented studies.

The studies presented in this report are based on three scenarios:

• Scenario 1: Year 2015, 2000MW installed wind generation capacity

• Scenario 2: Year 2020, 4800MW installed wind generation capacity

• Scenario 3: Year 2020, 10000MW installed wind generation capacity

For each of the three scenarios, the average loss of load probability (LOLP) at the daily peak load has been

calculated and has been used as the relevant reliability index for assessing the capacity credit of wind generation

in South Africa.

Capacity credit has been defined in terms of the Equivalent Firm Capacity (EFC), which is a generation capacity

with 100% availability that one would have to add to the system in order to achieve the same improvement in

terms of the available reserve at a given confidence level.

The resulting capacity credit figures for the different scenarios are the following:

• Scenario 1: CC=26,8%

• Scenario 2: CC=25,4%

• Scenario 3: CC=22,6%

With increasing wind penetration levels, the capacity credit of wind generation will drop. Based on Scenario 3

(year 2020, 10 000MW of installed wind capacity), an assessment has been carried out by scaling the 67

modelled wind farms up to a totally installed wind capacity of 25 000MW. The resulting capacity credit is equal to

17,6%.

Comparing these figures with the availability indices of newly planned coal fired power stations in South Africa

having a combined planned/unplanned outage rate of around 11%, the contribution of an average wind farm in

South Africa to the equivalent firm capacity of the system will be in the following range:

• Scenario 1: CC=29,8%

• Scenario 2: CC=26,1%

• Scenario 3: CC=25,1%

When comparing the capacity credit of wind generation to older coal fired power stations with a combined

planned/unplanned outage rate of around 20%, the capacity credit of wind generation will be around

33,5%/31,7%/28,2% (Sc1/Sc2/Sc3) of the capacity credit of such an older existing South African coal fired

power station.

Page 34: Giz capacity creditstudiessa_final

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 3 4

The second part of the presented studies analyses the impact of wind generation in South Africa on the residual

load, which is the remaining load that must be supplied by thermal and hydro power plants.

The analysis was mainly looking at:

• Worst case situations with regard to wind and load variations.

• Impact of wind generation on dynamic performance requirements.

The results of the corresponding studies lead to the conclusion that:

• Hourly ramp-up and ramp-down rates of the residual load are comparable to the corresponding ramp-

rates of the system load (without wind generation).

• There are no increased dynamic performance requirements for the existing thermal power plants in

South Africa.

The main impact on system operation will result from the limited predictability of wind speeds and not from

absolute wind speed variations. The limited predictability of wind generation will result in an increased forecast

error of the residual load compared to the present load forecast error. Several factors will have an influence on

the accuracy of wind prediction; some of them are related to the spatial distribution of wind generation, some of

them to the actual wind prediction system that will be put in place for supporting the operation of the South

African power system. Therefore, it is required to carry out additional studies that simulate the behaviour of a

wind prediction system in order to obtain indicative values for the required increase of the load following reserve.

As overall conclusion, it can be stated that the capacity credit of wind generation in South Africa will be between

25% and 30% for installed wind generation of up to 10 000MW. In the case of higher wind penetration (25

000MW), capacity credit of wind generation in South Africa will drop below 20%.

Based on part 2 of the presented studies it can further be concluded that it is very likely that it will be possible to

operate the system safely, without increased dynamic performance requirements for the conventional power

plants of South Africa. However, the use of state of the art wind prediction tools for assessing the required load

(and wind) following reserves will be important. This second aspects requires further, more detailed studies that

model the dynamic performance characteristics of the South African power plants in more detail for ensuring a

safe operation of the South African power system under all credible operating conditions.

Page 35: Giz capacity creditstudiessa_final

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 3 5

5 References

[1] GTZ: Kleinteilige Energie- und/oder Klimarelevante Maßnahmen – Teilwerk: Capacity Credit Study – Pr.-Nr.:

95.3550.1-034.00 – Terms of Reference, 01.06.2010

[2] ESKOM: Wind Project Sites for Study (Wind Project Sites for Study - 2010 May - Locations - MTS Sub - MW

sizes.xls), 2010

[3] ESKOM: File Wind_DigSilent_GxData_11062010.xlsx, 2010

[4] Windlab Systems: Capacity Credit Study in South Africa – Consulting Report, 14.07.2010

[5] Windlab Systems: Second version of wind speed data based on site adjustments, 15.11.2010

[6] ESKOM: Integrated Resource Plan for Electricity – Draft, Version 8, 08.10.2010 (IRP 2010)

[7] Windlab Systems: WindScapeTM Methodology – Capacity Credit Study in South Africa, 03.11.2010

Page 36: Giz capacity creditstudiessa_final

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 3 6

A N N E X E S

Annexes

Page 37: Giz capacity creditstudiessa_final

A n n e x 1 : R e s u l t s o f S i m u l a t i o n S t u d i e s

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 3 7

Annex 1: Results of Simulation Studies

Annex 1.1: Scenario 1 – Year 2015 – 2000MW Installed Wind Generation

Page 38: Giz capacity creditstudiessa_final

A n n e x 1 : R e s u l t s o f S i m u l a t i o n S t u d i e s

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 3 8

100,0080,00060,00040,00020,0000,0000 [%]

60000,00

50000,00

40000,00

30000,00

20000,00

10000,00

Summary Grid: Total Available Capacity in MW

Summary Grid: Available Dispatchable Capacity in MW

100,0080,00060,00040,00020,0000,0000 [%]

2.00E+4

1.50E+4

1.00E+4

5.00E+3

0.00E+0

-5.00E+3

Summary Grid: Total Reserve Generation (Unconstrained) in MW

Summary Grid: Reserve Dispatchable Generation (Unconstrained) in MW

100,0080,0060,0040,0020,000,00 [%]

60000,00

50000,00

40000,00

30000,00

20000,00

10000,00

Summary Grid: Total Demand in MW

Summary Grid: Residual Demand (Unconstrained) in MW

DIGSILENT Capacity Credit Studies for South Africa Distribution

Scenario 1 - 2015 - 2000MW installed Wind Capacity Winter season

Date: 12/25/2010

Annex: 1.1.1 /1

DIg

SIL

EN

T

Page 39: Giz capacity creditstudiessa_final

A n n e x 1 : R e s u l t s o f S i m u l a t i o n S t u d i e s

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 3 9

100,0099,8099,6099,4099,2099,00 [%]

8000,00

6000,00

4000,00

2000,00

0,00

-2000,00

Summary Grid: Total Reserve Generation (Unconstrained) in MW

Summary Grid: Reserve Dispatchable Generation (Unconstrained) in MW

X = 99,900 %

2251.571 MW

2786.828 MW

0.000 MW

Y =

0,0

00 M

W

99.998 %

DIGSILENT Capacity Credit Studies for South Africa Generation vs Demand

Scenario 1 - 2015 - 2000MW installed Wind Capacity Winter season

Date: 12/25/2010

Annex: 1.1.1 /2

DIg

SIL

EN

T

Page 40: Giz capacity creditstudiessa_final

A n n e x 1 : R e s u l t s o f S i m u l a t i o n S t u d i e s

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 4 0

Annex 1.2: Scenario 2 – Year 2020 – 4800MW Installed Wind Generation

Page 41: Giz capacity creditstudiessa_final

A n n e x 1 : R e s u l t s o f S i m u l a t i o n S t u d i e s

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 4 1

100,0080,00060,00040,00020,0000,0000 [%]

80000,00

70000,00

60000,00

50000,00

40000,00

30000,00

Summary Grid: Total Available Capacity in MW

Summary Grid: Available Dispatchable Capacity in MW

100,0080,0060,0040,0020,000,00 [%]

2.00E+4

1.50E+4

1.00E+4

5.00E+3

0.00E+0

-5.00E+3

Summary Grid: Total Reserve Generation (Unconstrained) in MW

Summary Grid: Reserve Dispatchable Generation (Unconstrained) in MW

100,0080,00060,00040,00020,0000,0000 [%]

80000,00

70000,00

60000,00

50000,00

40000,00

30000,00

Summary Grid: Daily Peak Load in MW

Summary Grid: Daily Peak of Residual Load in MW

DIGSILENT Capacity Credit Studies for South Africa Distribution

Scenario 2 - 2020/Low - 48000MW installed Wind Capacity Winter season

Date: 12/25/2010

Annex: 1.2.1 /1

DIg

SIL

EN

T

Page 42: Giz capacity creditstudiessa_final

A n n e x 1 : R e s u l t s o f S i m u l a t i o n S t u d i e s

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 4 2

100,0099,8099,6099,4099,2099,00 [%]

1.20E+4

9.00E+3

6.00E+3

3.00E+3

0.00E+0

-3.00E+3

Summary Grid: Total Reserve Generation (Unconstrained) in MW

Summary Grid: Reserve Dispatchable Generation (Unconstrained) in MW

X = 99,900 %

1459.123 MW

2677.937 MW

0.000 MWY = 0,000 MW99.991 %99.999 %

DIGSILENT Capacity Credit Studies for South Africa Generation vs. Demand

Scenario 2 - 2020/Low - 48000MW installed Wind Capacity Winter season

Date: 12/25/2010

Annex: 1.2.1 /2

DIg

SIL

EN

T

Page 43: Giz capacity creditstudiessa_final

A n n e x 1 : R e s u l t s o f S i m u l a t i o n S t u d i e s

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 4 3

Annex 1.3: Scenario 3 – Year 2020 – 10000MW Installed Wind Generation

Page 44: Giz capacity creditstudiessa_final

A n n e x 1 : R e s u l t s o f S i m u l a t i o n S t u d i e s

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 4 4

100,0080,00060,00040,00020,0000,0000 [%]

9.00E+4

7.00E+4

5.00E+4

3.00E+4

1.00E+4

-1.00E+4

Summary Grid: Total Available Capacity in MW

Summary Grid: Available Dispatchable Capacity in MW

100,0080,00060,00040,00020,0000,0000 [%]

9.00E+4

7.00E+4

5.00E+4

3.00E+4

1.00E+4

-1.00E+4

Summary Grid: Total Reserve Generation (Unconstrained) in MW

Summary Grid: Reserve Dispatchable Generation (Unconstrained) in MW

100,0080,0060,0040,0020,000,00 [%]

53000,00

48000,00

43000,00

38000,00

33000,00

28000,00

Summary Grid: Total Demand in MW

Summary Grid: Residual Demand (Unconstrained) in MW

0.291 %48316.363 MW

DIGSILENT Capacity Credit Studies for South Africa Distribution

Scenario 3 - 2020/High - 10000MW installed Wind Capacity Winter season

Date: 12/26/2010

Annex: 1.3.1 /1

DIg

SIL

EN

T

Page 45: Giz capacity creditstudiessa_final

A n n e x 1 : R e s u l t s o f S i m u l a t i o n S t u d i e s

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 4 5

100,0099,8099,6099,4099,2099,00 [%]

1.20E+4

9.00E+3

6.00E+3

3.00E+3

0.00E+0

-3.00E+3

Summary Grid: Total Reserve Generation (Unconstrained) in MW

Summary Grid: Reserve Dispatchable Generation (Unconstrained) in MW

X = 99,900 %

1455.504 MW

3715.627 MW

-0.000 MW

Y =

-0,0

00 M

W

99.993 %100.000 %

DIGSILENT Capacity Credit Studies for South Africa Generation vs Demand

Scenario 3 - 2020/High - 10000MW installed Wind Capacity Winter season

Date: 12/26/2010

Annex: 1.3.1 /2

DIg

SIL

EN

T

Page 46: Giz capacity creditstudiessa_final

A n n e x 2 : S e a s o n a l a n d H o u r l y V a r i a t i o n s o f t h e W i n d G e n e r a t i o n i n S o u t h A f r i c a

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 4 6

Annex 2: Seasonal and Hourly Variations of the Wind Generation in South Africa

There are the following dependencies of capacity credit on wind characteristics:

• Capacity credit increases with increasing average wind speeds.

• A high correlation between wind generation and load increases capacity credits.

With regard to correlation between wind generation and load, the two following aspects are of major importance:

• Correlation between seasonal wind variations and seasonal load variations.

• Correlation between hourly wind variations and hourly load variations.

For better understanding the results of the previous sections, the above correlation aspects have been analyzed

for the South African case.

Figure 18, Figure 21 and Figure 24 show the seasonal variations of wind generation in South Africa for the three

analyzed scenarios. These figures highlight the positive seasonal correlation between wind speeds and load in

South Africa.

Figure 19, Figure 22 and Figure 25 show the average wind generation per hour for the winter season and Figure

20, Figure 23 and Figure 26 for the summer season. Comparing these figures with the defined full load hours –

between 18:00h and 21:00h in winter and 08:00h and 21:00h in summer- it can be concluded that the correlation

between wind speeds and daily load variations is reasonably good:

• During winter, which is the most relevant case because this corresponds to the peak-load season, the

average wind generation during full load hours is approximately equal to the seasonal average.

• During summer, there is a very good correlation between wind generation and the evening load peak.

On the other hand, there is very low wind generation during morning hours, and hence only a low

contribution of wind generation to the morning high load coverage.

Page 47: Giz capacity creditstudiessa_final

A n n e x 2 : S e a s o n a l a n d H o u r l y V a r i a t i o n s o f t h e W i n d G e n e r a t i o n i n S o u t h A f r i c a

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0

Annex 2-1: Scenario 1 Capacity

Figure 18: Monthly Average of Wind Generation in South Africa

Figure 19: Hourly Average of Wind Generation in South Africa during winter

A n n e x 2 : S e a s o n a l a n d H o u r l y V a r i a t i o n s o f t h e W i n d G e n e r a t i o n i n S o u t h A f r i c a

0 3 4 . 0 0

1: Scenario 1 – Year 2015/2000MW of Installed Wind

: Monthly Average of Wind Generation in South Africa - Scenario 1

: Hourly Average of Wind Generation in South Africa during winter -

A n n e x 2 : S e a s o n a l a n d H o u r l y V a r i a t i o n s o f t h e W i n d G e n e r a t i o n i n S o u t h A f r i c a

4 7

Year 2015/2000MW of Installed Wind

- Scenario 1

Page 48: Giz capacity creditstudiessa_final

A n n e x 2 : S e a s o n a l a n d H o u r l y V a r i a t i o n s o f t h e W i n d G e n e r a t i o n i n S o u t h A f r i c a

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0

Figure 20: Hourly Average of Wind Generation in South Africa during summer

A n n e x 2 : S e a s o n a l a n d H o u r l y V a r i a t i o n s o f t h e W i n d G e n e r a t i o n i n S o u t h A f r i c a

0 3 4 . 0 0

: Hourly Average of Wind Generation in South Africa during summer

A n n e x 2 : S e a s o n a l a n d H o u r l y V a r i a t i o n s o f t h e W i n d G e n e r a t i o n i n S o u t h A f r i c a

4 8

: Hourly Average of Wind Generation in South Africa during summer - Scenario 1

Page 49: Giz capacity creditstudiessa_final

A n n e x 2 : S e a s o n a l a n d H o u r l y V a r i a t i o n s o f t h e W i n d G e n e r a t i o n i n S o u t h

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0

Annex 2-2: Scenario 2 Wind Capacity

Figure 21: Monthly Average of Wind Generation in South Africa

A n n e x 2 : S e a s o n a l a n d H o u r l y V a r i a t i o n s o f t h e W i n d G e n e r a t i o n i n S o u t h

0 3 4 . 0 0

2: Scenario 2 – Year 2020-Low/4800MW of Installed

: Monthly Average of Wind Generation in South Africa - Scenario 2

A n n e x 2 : S e a s o n a l a n d H o u r l y V a r i a t i o n s o f t h e W i n d G e n e r a t i o n i n S o u t h A f r i c a

4 9

w/4800MW of Installed

Page 50: Giz capacity creditstudiessa_final

A n n e x 2 : S e a s o n a l a n d H o u r l y V a r i a t i o n s o f t h e W i n d G e n e r a t i o n i n S o u t h A f r i c a

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0

Figure 22: Hourly Average of Wind Generation in South Africa during winter

Figure 23: Hourly Average of Wind Generation in South Africa during summer

A n n e x 2 : S e a s o n a l a n d H o u r l y V a r i a t i o n s o f t h e W i n d G e n e r a t i o n i n S o u t h A f r i c a

0 3 4 . 0 0

: Hourly Average of Wind Generation in South Africa during winter -

: Hourly Average of Wind Generation in South Africa during summer

A n n e x 2 : S e a s o n a l a n d H o u r l y V a r i a t i o n s o f t h e W i n d G e n e r a t i o n i n S o u t h A f r i c a

5 0

- Scenario 2

: Hourly Average of Wind Generation in South Africa during summer - Scenario 2

Page 51: Giz capacity creditstudiessa_final

A n n e x 2 : S e a s o n a l a n d H o u r l y V a r i a t i o n s o f t h e W i n d G e n e r a t i o n i n S o u t h A f r i c a

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0

Annex 2-3: Scenario 3 Wind Capacity

Figure 24: Monthly Average of Wind

A n n e x 2 : S e a s o n a l a n d H o u r l y V a r i a t i o n s o f t h e W i n d G e n e r a t i o n i n S o u t h A f r i c a

0 3 4 . 0 0

3: Scenario 3 – Year 2020-High/10000MW of Installed

Monthly Average of Wind Generation in South Africa - Scenario 3

A n n e x 2 : S e a s o n a l a n d H o u r l y V a r i a t i o n s o f t h e W i n d G e n e r a t i o n i n S o u t h A f r i c a

5 1

High/10000MW of Installed

Page 52: Giz capacity creditstudiessa_final

A n n e x 2 : S e a s o n a l a n d H o u r l y V a r i a t i o n s o f t h e W i n d G e n e r a t i o n i n S o u t h A f r i c a

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0

Figure 25: Hourly Average of Wind Generation in South Africa during winter

Figure 26: Hourly Average of Wind Generation in South Africa during summe

A n n e x 2 : S e a s o n a l a n d H o u r l y V a r i a t i o n s o f t h e W i n d G e n e r a t i o n i n S o u t h A f r i c a

0 3 4 . 0 0

Hourly Average of Wind Generation in South Africa during winter -

: Hourly Average of Wind Generation in South Africa during summe

A n n e x 2 : S e a s o n a l a n d H o u r l y V a r i a t i o n s o f t h e W i n d G e n e r a t i o n i n S o u t h A f r i c a

5 2

- Scenario 3

: Hourly Average of Wind Generation in South Africa during summer - Scenario 3

Page 53: Giz capacity creditstudiessa_final

A n n e x 3 : R e s u l t s o f T i m e S e r i e s A s s e s s m e n t

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 5 3

Annex 3: Results of Time Series Assessment

Annex 3-1: Worst Case Situations and Duration Curves, Scenario 1 – 2015

Page 54: Giz capacity creditstudiessa_final

A n n e x 3 : R e s u l t s o f T i m e S e r i e s A s s e s s m e n t

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 5 4

5000,04700,04400,04100,03800,03500,0

2000,0

1600,0

1200,0

800,00

400,00

0,0000

x-Axis: TimeSeries: Hour of year in h

TimeSeries: Total Wind Generation in MW

5000,04700,04400,04100,03800,03500,0

5.00E+4

3.75E+4

2.50E+4

1.25E+4

0.00E+0

-1.25E+4

x-Axis: TimeSeries: Hour of year in h

TimeSeries: Total Load in MW

TimeSeries: Total Residual Load in MW

TimeSeries: Total Wind Generation in MW

DIgSILENT Capacity Credit Studies for South Africa - Part 2 LongerTerm_Winter

Time Series Studies - Scenario 1 Worst Case Situations

Date: 12/26/2010

Annex: Annex 3.1 /1

DIg

SIL

EN

T

Page 55: Giz capacity creditstudiessa_final

A n n e x 3 : R e s u l t s o f T i m e S e r i e s A s s e s s m e n t

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 5 5

8700,08360,08020,07680,07340,07000,0

2000,0

1600,0

1200,0

800,00

400,00

0,0000

x-Axis: TimeSeries: Hour of year in h

TimeSeries: Total Wind Generation in MW

8700,08360,08020,07680,07340,07000,0

5.00E+4

3.75E+4

2.50E+4

1.25E+4

0.00E+0

-1.25E+4

x-Axis: TimeSeries: Hour of year in h

TimeSeries: Total Load in MW

TimeSeries: Total Residual Load in MW

TimeSeries: Total Wind Generation in MW

DIgSILENT Capacity Credit Studies for South Africa - Part 2 LongerTerm_Summer

Time Series Studies - Scenario 1 Worst Case Situations

Date: 12/26/2010

Annex: Annex 3.1 /2

DIg

SIL

EN

T

Page 56: Giz capacity creditstudiessa_final

A n n e x 3 : R e s u l t s o f T i m e S e r i e s A s s e s s m e n t

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 5 6

4218,04184,44150,84117,24083,64050,0

2000,0

1600,0

1200,0

800,00

400,00

0,0000

x-Axis: TimeSeries: Hour of year in h

TimeSeries: Total Wind Generation in MW

4136.000 4136.000 h1738.038 MW

4218,04184,44150,84117,24083,64050,0

5.00E+4

3.75E+4

2.50E+4

1.25E+4

0.00E+0

-1.25E+4

x-Axis: TimeSeries: Hour of year in h

TimeSeries: Total Load in MW

TimeSeries: Total Residual Load in MW

TimeSeries: Total Wind Generation in MW

DIgSILENT Capacity Credit Studies for South Africa - Part 2 MaxWind

Time Series Studies - Scenario 1 Worst Case Situations

Date: 12/26/2010

Annex: Annex 3.1 /3

DIg

SIL

EN

T

Page 57: Giz capacity creditstudiessa_final

A n n e x 3 : R e s u l t s o f T i m e S e r i e s A s s e s s m e n t

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 5 7

7168,07134,47100,87067,27033,67000,0

2000,0

1600,0

1200,0

800,00

400,00

0,0000

x-Axis: TimeSeries: Hour of year in h

TimeSeries: Total Wind Generation in MW

7066.000 7066.000 h10.286 MW

7168,07134,47100,87067,27033,67000,0

5.00E+4

3.75E+4

2.50E+4

1.25E+4

0.00E+0

-1.25E+4

x-Axis: TimeSeries: Hour of year in h

TimeSeries: Total Load in MW

TimeSeries: Total Residual Load in MW

TimeSeries: Total Wind Generation in MW

DIgSILENT Capacity Credit Studies for South Africa - Part 2 MinWind

Time Series Studies - Scenario 1 Worst Case Situations

Date: 12/26/2010

Annex: Annex 3.1 /4

DIg

SIL

EN

T

Page 58: Giz capacity creditstudiessa_final

A n n e x 3 : R e s u l t s o f T i m e S e r i e s A s s e s s m e n t

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 5 8

8568,08534,48500,88467,28433,68400,0

2000,0

1600,0

1200,0

800,00

400,00

0,0000

x-Axis: TimeSeries: Hour of year in h

TimeSeries: Total Wind Generation in MW

Y =502,404*MW/h*x-4276921,668 MW

8568,08534,48500,88467,28433,68400,0

5.00E+4

3.75E+4

2.50E+4

1.25E+4

0.00E+0

-1.25E+4

x-Axis: TimeSeries: Hour of year in h

TimeSeries: Total Load in MW

TimeSeries: Total Residual Load in MW

TimeSeries: Total Wind Generation in MW

DIgSILENT Capacity Credit Studies for South Africa - Part 2 MaxWindUp

Time Series Studies - Scenario 1 Worst Case Situations

Date: 12/26/2010

Annex: Annex 3.1 /5

DIg

SIL

EN

T

Page 59: Giz capacity creditstudiessa_final

A n n e x 3 : R e s u l t s o f T i m e S e r i e s A s s e s s m e n t

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 5 9

168,00134,40100,8067,20033,6000,0000 [-]

2000,0

1600,0

1200,0

800,00

400,00

0,0000

TimeSeries: Total Wind Generation in MW

Y =

-355,9

99*M

W/

*x+

1193,2

20 M

W

168,00134,40100,8067,20033,6000,0000 [-]

4.00E+4

3.00E+4

2.00E+4

1.00E+4

0.00E+0

-1.00E+4

TimeSeries: Total Load in MW

TimeSeries: Total Residual Load in MW

TimeSeries: Total Wind Generation in MW

DIgSILENT Capacity Credit Studies for South Africa - Part 2 MaxWindDown

Time Series Studies - Scenario 1 Worst Case Situations

Date: 12/26/2010

Annex: Annex 3.1 /6

DIg

SIL

EN

T

Page 60: Giz capacity creditstudiessa_final

A n n e x 3 : R e s u l t s o f T i m e S e r i e s A s s e s s m e n t

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 6 0

8618,08584,48550,88517,28483,68450,0

2000,0

1600,0

1200,0

800,00

400,00

0,0000

x-Axis: TimeSeries: Hour of year in h

TimeSeries: Total Wind Generation in MW

8618,08584,48550,88517,28483,68450,0

4.00E+4

3.00E+4

2.00E+4

1.00E+4

0.00E+0

-1.00E+4

x-Axis: TimeSeries: Hour of year in h

TimeSeries: Total Load in MW

TimeSeries: Total Residual Load in MW

TimeSeries: Total Wind Generation in MW

8525.000 8525.000 h19738.392 MW

DIgSILENT Capacity Credit Studies for South Africa - Part 2 MinResLoad

Time Series Studies - Scenario 1 Worst Case Situations

Date: 12/26/2010

Annex: Annex 3.1 /7

DIg

SIL

EN

T

Page 61: Giz capacity creditstudiessa_final

A n n e x 3 : R e s u l t s o f T i m e S e r i e s A s s e s s m e n t

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 6 1

4468,04434,44400,84367,24333,64300,0

2000,0

1600,0

1200,0

800,00

400,00

0,0000

x-Axis: TimeSeries: Hour of year in h

TimeSeries: Total Wind Generation in MW

4468,04434,44400,84367,24333,64300,0

5.00E+4

3.75E+4

2.50E+4

1.25E+4

0.00E+0

-1.25E+4

x-Axis: TimeSeries: Hour of year in h

TimeSeries: Total Load in MW

TimeSeries: Total Residual Load in MW

TimeSeries: Total Wind Generation in MW

4388.000 4388.000 h40581.754 MW

DIgSILENT Capacity Credit Studies for South Africa - Part 2 MaxLoad

Time Series Studies - Scenario 1 Worst Case Situations

Date: 12/26/2010

Annex: Annex 3.1 /8

DIg

SIL

EN

T

Page 62: Giz capacity creditstudiessa_final

A n n e x 3 : R e s u l t s o f T i m e S e r i e s A s s e s s m e n t

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 6 2

100,0080,00060,00040,00020,0000,0000

2000,00

1500,00

1000,00

500,00

0,00

-500,00

x-Axis: TimeSeries: Proability (Duration Curves) in %

TimeSeries: Wind Generation Duration Curve in MW

100,0080,00060,00040,00020,0000,0000

43000,00

38000,00

33000,00

28000,00

23000,00

18000,00

x-Axis: TimeSeries: Proability (Duration Curves) in %

TimeSeries: Load Duration Curve in MW

TimeSeries: Residual Load Duration Curve n MW

DIgSILENT Capacity Credit Studies for South Africa - Part 2 Duration

Time Series Studies - Scenario 1 Worst Case Situations

Date: 12/26/2010

Annex: Annex 3.1 /9

DIg

SIL

EN

T

Page 63: Giz capacity creditstudiessa_final

A n n e x 3 : R e s u l t s o f T i m e S e r i e s A s s e s s m e n t

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 6 3

25,0020,0015,0010,005,000,00

7500,00

5000,00

2500,00

0,00

-2500,00

-5000,00

x-Axis: TimeSeries: Proability (Duration Curves) in %

TimeSeries: Hourly Load Variation (up) in MW/h

TimeSeries: Hourly Load Variation (down) in MW/h

TimeSeries: Hourly Wind Variation (up) in MW/h

TimeSeries: Hourly Wind Variation (down) in MW/h

TimeSeries: Hourly Residual Load Variation (up) in MW/h

TimeSeries: Hourly Residual Load Variation (down) in MW/h

DIgSILENT Capacity Credit Studies for South Africa - Part 2 Duration_Variations

Time Series Studies - Scenario 1 Worst Case Situations

Date: 12/26/2010

Annex: Annex 3.1 /10

DIg

SIL

EN

T

Page 64: Giz capacity creditstudiessa_final

A n n e x 3 : R e s u l t s o f T i m e S e r i e s A s s e s s m e n t

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 6 4

Annex 3-2: Worst Case Situations and Duration Curves, Scenario 2 – 2020

Page 65: Giz capacity creditstudiessa_final

A n n e x 3 : R e s u l t s o f T i m e S e r i e s A s s e s s m e n t

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 6 5

5000,04700,04400,04100,03800,03500,0 [-]

4800,0

3840,0

2880,0

1920,0

960,00

0,0000

TimeSeries: Total Wind Generation in MW

5000,04700,04400,04100,03800,03500,0 [-]

50000,

40000,

30000,

20000,

10000,

0,0000

TimeSeries: Total Load in MW

TimeSeries: Total Residual Load in MW

TimeSeries: Total Wind Generation in MW

DIgSILENT Capacity Credit Studies for South Africa - Part 2 LongerTerm_Winter

Time Series Studies - Scenario 2 Worst Case Situations

Date: 12/26/2010

Annex: Annex 3.2 /1

DIg

SIL

EN

T

Page 66: Giz capacity creditstudiessa_final

A n n e x 3 : R e s u l t s o f T i m e S e r i e s A s s e s s m e n t

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 6 6

8700,08360,08020,07680,07340,07000,0 [-]

4800,0

3840,0

2880,0

1920,0

960,00

0,0000

TimeSeries: Total Wind Generation in MW

8700,08360,08020,07680,07340,07000,0 [-]

50000,

40000,

30000,

20000,

10000,

0,0000

TimeSeries: Total Load in MW

TimeSeries: Total Residual Load in MW

TimeSeries: Total Wind Generation in MW

DIgSILENT Capacity Credit Studies for South Africa - Part 2 LongerTerm_Summer

Time Series Studies - Scenario 2 Worst Case Situations

Date: 12/26/2010

Annex: Annex 3.2 /2

DIg

SIL

EN

T

Page 67: Giz capacity creditstudiessa_final

A n n e x 3 : R e s u l t s o f T i m e S e r i e s A s s e s s m e n t

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 6 7

4768,04734,44700,84667,24633,64600,0 [-]

4800,0

3844,1

2888,2

1932,2

976,31

20,388

TimeSeries: Total Wind Generation in MW

4687.000 4226.631 MW

4768,04734,44700,84667,24633,64600,0 [-]

5.04E+4

3.99E+4

2.94E+4

1.88E+4

8.33E+3

-2.18E+3

TimeSeries: Total Load in MW

TimeSeries: Total Residual Load in MW

TimeSeries: Total Wind Generation in MW

DIgSILENT Capacity Credit Studies for South Africa - Part 2 MaxWind

Time Series Studies - Scenario 2 Worst Case Situations

Date: 12/26/2010

Annex: Annex 3.2 /3

DIg

SIL

EN

T

Page 68: Giz capacity creditstudiessa_final

A n n e x 3 : R e s u l t s o f T i m e S e r i e s A s s e s s m e n t

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 6 8

8168,08134,48100,88067,28033,68000,0 [-]

4800,0

3840,0

2880,0

1920,0

960,00

0,0000

TimeSeries: Total Wind Generation in MW

8048.000 26.425 MW

8168,08134,48100,88067,28033,68000,0 [-]

50000,

40000,

30000,

20000,

10000,

0,0000

TimeSeries: Total Load in MW

TimeSeries: Total Residual Load in MW

TimeSeries: Total Wind Generation in MW

DIgSILENT Capacity Credit Studies for South Africa - Part 2 MinWind

Time Series Studies - Scenario 2 Worst Case Situations

Date: 12/26/2010

Annex: Annex 3.2 /4

DIg

SIL

EN

T

Page 69: Giz capacity creditstudiessa_final

A n n e x 3 : R e s u l t s o f T i m e S e r i e s A s s e s s m e n t

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 6 9

168,00134,40100,8067,20033,6000,0000 [-]

4800,0

3840,0

2880,0

1920,0

960,00

0,0000

TimeSeries: Total Wind Generation in MW

Y =

-1101,7

48*M

W/

*x+

2787,2

25 M

W

168,00134,40100,8067,20033,6000,0000 [-]

50000,

40000,

30000,

20000,

10000,

0,0000

TimeSeries: Total Load in MW

TimeSeries: Total Residual Load in MW

TimeSeries: Total Wind Generation in MW

DIgSILENT Capacity Credit Studies for South Africa - Part 2 MaxWindDown

Time Series Studies - Scenario 2 Worst Case Situations

Date: 12/26/2010

Annex: Annex 3.2 /5

DIg

SIL

EN

T

Page 70: Giz capacity creditstudiessa_final

A n n e x 3 : R e s u l t s o f T i m e S e r i e s A s s e s s m e n t

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 7 0

960,00928,00896,00864,00832,00800,00 [-]

4800,0

3840,0

2880,0

1920,0

960,00

0,0000

TimeSeries: Total Wind Generation in MW

Y =1369,078*MW/ *x-1171796,607 MW

960,00928,00896,00864,00832,00800,00 [-]

50000,

40000,

30000,

20000,

10000,

0,0000

TimeSeries: Total Load in MW

TimeSeries: Total Residual Load in MW

TimeSeries: Total Wind Generation in MW

DIgSILENT Capacity Credit Studies for South Africa - Part 2 MaxWindUp

Time Series Studies - Scenario 2 Worst Case Situations

Date: 12/26/2010

Annex: Annex 3.2 /6

DIg

SIL

EN

T

Page 71: Giz capacity creditstudiessa_final

A n n e x 3 : R e s u l t s o f T i m e S e r i e s A s s e s s m e n t

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 7 1

8760,08726,48692,88659,28625,68592,0 [-]

5000,00

4000,00

3000,00

2000,00

1000,00

0,00

TimeSeries: Total Wind Generation in MW

8760,08726,48692,88659,28625,68592,0 [-]

50000,00

40000,00

30000,00

20000,00

10000,00

0,00

TimeSeries: Total Load in MW

TimeSeries: Total Residual Load in MW

TimeSeries: Total Wind Generation in MW

8717.000 22534.721 MW

DIgSILENT Capacity Credit Studies for South Africa - Part 2 MinResLoad

Time Series Studies - Scenario 2 Worst Case Situations

Date: 12/26/2010

Annex: Annex 3.2 /7

DIg

SIL

EN

T

Page 72: Giz capacity creditstudiessa_final

A n n e x 3 : R e s u l t s o f T i m e S e r i e s A s s e s s m e n t

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 7 2

4468,04434,44400,84367,24333,64300,0 [-]

5000,00

4000,00

3000,00

2000,00

1000,00

0,00

TimeSeries: Total Wind Generation in MW

4468,04434,44400,84367,24333,64300,0 [-]

50000,

40000,

30000,

20000,

10000,

0,0000

TimeSeries: Total Load in MW

TimeSeries: Total Residual Load in MW

TimeSeries: Total Wind Generation in MW

4388.000 48316.363 MW

DIgSILENT Capacity Credit Studies for South Africa - Part 2 MaxLoad

Time Series Studies - Scenario 2 Worst Case Situations

Date: 12/26/2010

Annex: Annex 3.2 /8

DIg

SIL

EN

T

Page 73: Giz capacity creditstudiessa_final

A n n e x 3 : R e s u l t s o f T i m e S e r i e s A s s e s s m e n t

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 7 3

100,0080,00060,00040,00020,0000,0000

5000,00

3750,00

2500,00

1250,00

0,00

-1250,00

x-Axis: TimeSeries: Proability (Duration Curves) in %

TimeSeries: Wind Generation Duration Curve in MW

100,0080,00060,00040,00020,0000,0000

70000,00

60000,00

50000,00

40000,00

30000,00

20000,00

x-Axis: TimeSeries: Proability (Duration Curves) in %

TimeSeries: Load Duration Curve in MW

TimeSeries: Residual Load Duration Curve n MW

DIgSILENT Capacity Credit Studies for South Africa - Part 2 Duration

Time Series Studies - Scenario 2 Worst Case Situations

Date: 12/26/2010

Annex: Annex 3.2 /9

DIg

SIL

EN

T

Page 74: Giz capacity creditstudiessa_final

A n n e x 3 : R e s u l t s o f T i m e S e r i e s A s s e s s m e n t

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 7 4

25,0020,0015,0010,005,000,00

9000,00

6000,00

3000,00

0,00

-3000,00

-6000,00

x-Axis: TimeSeries: Proability (Duration Curves) in %

TimeSeries: Hourly Load Variation (up) in MW/h

TimeSeries: Hourly Load Variation (down) in MW/h

TimeSeries: Hourly Wind Variation (up) in MW/h

TimeSeries: Hourly Wind Variation (down) in MW/h

TimeSeries: Hourly Residual Load Variation (up) in MW/h

TimeSeries: Hourly Residual Load Variation (down) in MW/h

DIgSILENT Capacity Credit Studies for South Africa - Part 2 Duration_Variations

Time Series Studies - Scenario 2 Worst Case Situations

Date: 12/26/2010

Annex: Annex 3.2 /10

DIg

SIL

EN

T

Page 75: Giz capacity creditstudiessa_final

A n n e x 3 : R e s u l t s o f T i m e S e r i e s A s s e s s m e n t

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 7 5

Annex 3-3: Worst Case Situations and Duration Curves, Scenario 3 – 2020

Page 76: Giz capacity creditstudiessa_final

A n n e x 3 : R e s u l t s o f T i m e S e r i e s A s s e s s m e n t

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 7 6

5000,04700,04400,04100,03800,03500,0

10000,00

8000,00

6000,00

4000,00

2000,00

0,00

x-Axis: TimeSeries: Hour of year in h

TimeSeries: Total Wind Generation in MW

5000,04700,04400,04100,03800,03500,0

50000,

40000,

30000,

20000,

10000,

0,0000

x-Axis: TimeSeries: Hour of year in h

TimeSeries: Total Load in MW

TimeSeries: Total Residual Load in MW

TimeSeries: Total Wind Generation in MW

DIgSILENT Capacity Credit Studies for South Africa - Part 2 LongerTerm_Winter

Time Series Studies - Scenario 3 Worst Case Situations

Date: 12/26/2010

Annex: Annex 3.3 /1

DIg

SIL

EN

T

Page 77: Giz capacity creditstudiessa_final

A n n e x 3 : R e s u l t s o f T i m e S e r i e s A s s e s s m e n t

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 7 7

8700,08360,08020,07680,07340,07000,0

10000,

8000,0

6000,0

4000,0

2000,0

0,0000

x-Axis: TimeSeries: Hour of year in h

TimeSeries: Total Wind Generation in MW

8700,08360,08020,07680,07340,07000,0

50000,00

40000,00

30000,00

20000,00

10000,00

0,00

x-Axis: TimeSeries: Hour of year in h

TimeSeries: Total Load in MW

TimeSeries: Total Residual Load in MW

TimeSeries: Total Wind Generation in MW

DIgSILENT Capacity Credit Studies for South Africa - Part 2 LongerTerm_Summer

Time Series Studies - Scenario 3 Worst Case Situations

Date: 12/26/2010

Annex: Annex 3.3 /2

DIg

SIL

EN

T

Page 78: Giz capacity creditstudiessa_final

A n n e x 3 : R e s u l t s o f T i m e S e r i e s A s s e s s m e n t

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 7 8

5168,05134,45100,85067,25033,65000,0

10000,00

8000,00

6000,00

4000,00

2000,00

0,00

x-Axis: TimeSeries: Hour of year in h

TimeSeries: Total Wind Generation in MW

5053.000 5053.000 h8468.058 MW

5168,05134,45100,85067,25033,65000,0

50000,

40000,

30000,

20000,

10000,

0,0000

x-Axis: TimeSeries: Hour of year in h

TimeSeries: Total Load in MW

TimeSeries: Total Residual Load in MW

TimeSeries: Total Wind Generation in MW

DIgSILENT Capacity Credit Studies for South Africa - Part 2 MaxWind

Time Series Studies - Scenario 3 Worst Case Situations

Date: 12/26/2010

Annex: Annex 3.3 /3

DIg

SIL

EN

T

Page 79: Giz capacity creditstudiessa_final

A n n e x 3 : R e s u l t s o f T i m e S e r i e s A s s e s s m e n t

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 7 9

8168,08134,48100,88067,28033,68000,0

10000,00

8000,00

6000,00

4000,00

2000,00

0,00

x-Axis: TimeSeries: Hour of year in h

TimeSeries: Total Wind Generation in MW

8048.000 8048.000 h78.456 MW

8168,08134,48100,88067,28033,68000,0

50000,00

40000,00

30000,00

20000,00

10000,00

0,00

x-Axis: TimeSeries: Hour of year in h

TimeSeries: Total Load in MW

TimeSeries: Total Residual Load in MW

TimeSeries: Total Wind Generation in MW

DIgSILENT Capacity Credit Studies for South Africa - Part 2 MinWind

Time Series Studies - Scenario 3 Worst Case Situations

Date: 12/26/2010

Annex: Annex 3.3 /4

DIg

SIL

EN

T

Page 80: Giz capacity creditstudiessa_final

A n n e x 3 : R e s u l t s o f T i m e S e r i e s A s s e s s m e n t

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 8 0

168,00134,40100,8067,20033,6000,0000

10000,00

8000,00

6000,00

4000,00

2000,00

0,00

x-Axis: TimeSeries: Hour of year in h

TimeSeries: Total Wind Generation in MW

Y =

-1670,4

81*M

W/h

*x+

5818,2

05 M

W

168,00134,40100,8067,20033,6000,0000

50000,

40000,

30000,

20000,

10000,

0,0000

x-Axis: TimeSeries: Hour of year in h

TimeSeries: Total Load in MW

TimeSeries: Total Residual Load in MW

TimeSeries: Total Wind Generation in MW

DIgSILENT Capacity Credit Studies for South Africa - Part 2 MaxWindDown

Time Series Studies - Scenario 3 Worst Case Situations

Date: 12/26/2010

Annex: Annex 3.3 /5

DIg

SIL

EN

T

Page 81: Giz capacity creditstudiessa_final

A n n e x 3 : R e s u l t s o f T i m e S e r i e s A s s e s s m e n t

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 8 1

968,00934,40900,80867,20833,60800,00

10000,00

8000,00

6000,00

4000,00

2000,00

0,00

x-Axis: TimeSeries: Hour of year in h

TimeSeries: Total Wind Generation in MW

Y =2546,982*MW/h*x-2179586,034 MW

968,00934,40900,80867,20833,60800,00

50000,

40000,

30000,

20000,

10000,

0,0000

x-Axis: TimeSeries: Hour of year in h

TimeSeries: Total Load in MW

TimeSeries: Total Residual Load in MW

TimeSeries: Total Wind Generation in MW

DIgSILENT Capacity Credit Studies for South Africa - Part 2 MaxWindUp

Time Series Studies - Scenario 3 Worst Case Situations

Date: 12/26/2010

Annex: Annex 3.3 /6

DIg

SIL

EN

T

Page 82: Giz capacity creditstudiessa_final

A n n e x 3 : R e s u l t s o f T i m e S e r i e s A s s e s s m e n t

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 8 2

8760,08726,48692,88659,28625,68592,0

10000,

8000,0

6000,0

4000,0

2000,0

0,0000

x-Axis: TimeSeries: Hour of year in h

TimeSeries: Total Wind Generation in MW

8760,08726,48692,88659,28625,68592,0

50000,

40000,

30000,

20000,

10000,

0,0000

x-Axis: TimeSeries: Hour of year in h

TimeSeries: Total Load in MW

TimeSeries: Total Residual Load in MW

TimeSeries: Total Wind Generation in MW

8717.000 8717.000 h21383.508 MW

DIgSILENT Capacity Credit Studies for South Africa - Part 2 MinResLoad

Time Series Studies - Scenario 3 Worst Case Situations

Date: 12/26/2010

Annex: Annex 3.3 /7

DIg

SIL

EN

T

Page 83: Giz capacity creditstudiessa_final

A n n e x 3 : R e s u l t s o f T i m e S e r i e s A s s e s s m e n t

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 8 3

4468,04434,44400,84367,24333,64300,0

10000,

8000,0

6000,0

4000,0

2000,0

0,0000

x-Axis: TimeSeries: Hour of year in h

TimeSeries: Total Wind Generation in MW

4468,04434,44400,84367,24333,64300,0

50000,

40000,

30000,

20000,

10000,

0,0000

x-Axis: TimeSeries: Hour of year in h

TimeSeries: Total Load in MW

TimeSeries: Total Residual Load in MW

TimeSeries: Total Wind Generation in MW

4388.000 4388.000 h48316.363 MW

DIgSILENT Capacity Credit Studies for South Africa - Part 2 MaxLoad

Time Series Studies - Scenario 3 Worst Case Situations

Date: 12/26/2010

Annex: Annex 3.3 /8

DIg

SIL

EN

T

Page 84: Giz capacity creditstudiessa_final

A n n e x 3 : R e s u l t s o f T i m e S e r i e s A s s e s s m e n t

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 8 4

100,0080,00060,00040,00020,0000,0000

1.00E+4

7.50E+3

5.00E+3

2.50E+3

0.00E+0

-2.50E+3

x-Axis: TimeSeries: Proability (Duration Curves) in %

TimeSeries: Wind Generation Duration Curve in MW

100,0080,00060,00040,00020,0000,0000

60000,00

50000,00

40000,00

30000,00

20000,00

10000,00

x-Axis: TimeSeries: Proability (Duration Curves) in %

TimeSeries: Load Duration Curve in MW

TimeSeries: Residual Load Duration Curve n MW

DIgSILENT Capacity Credit Studies for South Africa - Part 2 Duration

Time Series Studies - Scenario 3 Worst Case Situations

Date: 12/26/2010

Annex: Annex 3.3 /9

DIg

SIL

EN

T

Page 85: Giz capacity creditstudiessa_final

A n n e x 3 : R e s u l t s o f T i m e S e r i e s A s s e s s m e n t

P 1 4 2 9 - P N : 9 5 . 3 5 5 0 . 1 - 0 3 4 . 0 0 8 5

25,0020,0015,0010,005,000,00

9000,00

6000,00

3000,00

0,00

-3000,00

-6000,00

x-Axis: TimeSeries: Proability (Duration Curves) in %

TimeSeries: Hourly Load Variation (up) in MW/h

TimeSeries: Hourly Load Variation (down) in MW/h

TimeSeries: Hourly Wind Variation (up) in MW/h

TimeSeries: Hourly Wind Variation (down) in MW/h

TimeSeries: Hourly Residual Load Variation (up) in MW/h

TimeSeries: Hourly Residual Load Variation (down) in MW/h

DIgSILENT Capacity Credit Studies for South Africa - Part 2 Duration_Variations

Time Series Studies - Scenario 3 Worst Case Situations

Date: 12/26/2010

Annex: Annex 3.3 /10

DIg

SIL

EN

T