This preview shows page 1. Sign up to view the full content.
Unformatted text preview: ower stations can either
supply electricity directly to customers, or through the transmission network. 39 In South Africa a number of municipalities and large customers have their own
power stations. The enduse load profiles can change significantly, depending on how those power stations are operated. The utility’s generation pattern (a schedule of the maximum power station
capacities and imports from neighbouring countries available in the following
years) have to be studied to ensure the power stations are included in the
data files for power flow studies. Furthermore, a number of other aspects
have to be considered. The availability of water to hydroelectric power stations depends on droughts and irrigation quotas. Low fuel reserves at power stations can also reduce the expected net sent outs. It also has to be
ensured that the new power stations planned for the future are modelled and
the old power stations to be decommissioned are excluded.
The different generator base loads (minimum generating conditions  based
technical specifications and design criteria) have to be documented.
The net sent from each unit is determined by:
Net Sent Out = Base + (Max − Base)* p (3.2.3) where
Base minimum generating load depending on the
technical specifications and design criteria. Max maximum unit rating p A percentage depending on the generation cost From equation (3.2.3), when p equals 0 %, then the net pow er output equals
the base load and when p equals 100 %, then the net power output equals the
maximum unit rating. In most cases power stations with high generation costs
have lower p values. For a power station, equation (3.2.3) becomes
Net Sent Out = number of units ∑
i =1 Basei + (Maxi − Basei )* pi (3.2.4) 40 Network constraints have limitations on the net power outputs from power
stations and therefore it is important that all network topology changes are
correctly modelled. In the early stages of the research, chaos theory was been studied to predict
the generation pattern. The study was stopped, because it is not required for
network studies as mentioned above. Chaos theory is about explaining apparent disorder in a very ordered way.
Chaos theory states that things are not really random, just complex. Some
events that look random can be represented by a simple computation. After a
number of iterations, the simple computation will then produce the required
complex results. [12 – 16]
3.2.3 Power Stations
Any power station requires some auxiliary services to generate electricity.
Some of those auxiliary services are located outside the power station’s main
building. The electrical auxiliary supplies are normally connected to the station’s electrical boards, which are again electrically connected to the unit
boards inside the power station’s main building. In most cases the station
transformer supplies all the electricity required by the station’s electrical
boards. The high voltage side of the station transformer is either connected to
a distribution substation, or to a transmission substation.
The power station loads supplied from the transmission network are classified
as transmission loads. The utility has two pumpedstorage power stations. A
pumpedstorage power station generates during peak conditions (as hydro electric power station) and pumps water to its storage dams during offpeak
conditions. The two pumped storage power stations generate at the time of
maximum system demand and will therefore not take the pumping loads into
account.
The total power station loads are small compared to the system peak of 27
MW (see Chapter 5). 41 3.2.4 Transmission Losses
The electrical losses in an electrical network is defined as P loss = I2R where I is
the root mean square (rms) current in the conductors and R is the ac
resistance of the conductors. The losses are modelled as 3.2 % of the maximum system load see (Chapter 5). [17] To determine the exact electrical losses will complicate the balancing
algorithm and slow down the computational time. The figure assumed for
electrical losses can change for different generation patterns, or large loads
commissioned far from the power pool, or when new transmission lines are
built, etc. The question that always remains is what is more important: a feasible
solution (not the best, but acceptable for the purpose), or an optimal solution
(the best but maybe long computational times). Computational time is of essence as long as the results are feasible. This principle is followed throughout this chapter and in Chapter 4. The rule of thumb (3.2 % of the
system peak) is therefore acceptable and will be adjusted if necessary.
3.2.5 International Customers
The enduse customers outside the South African borders are called
Internatio nal Customers. Although small compared to the maximum system
load, approximately four percent (Chapter 5) are very difficult to determine
and to allocate to a supply point in the transmission network.
For example, the power flow on a 400 kV line can be...
View
Full
Document
This document was uploaded on 03/04/2014.
 Spring '14

Click to edit the document details