Don Geddey has over 40 years’ engineering experience in power system analysis and planning. Don discusses the importance of stable system frequency and why it’s becoming more challenging to maintain.
A healthy frequency range of 49.9 – 50.1 Hertz
Power systems aim to provide customers with voltages of a near-constant magnitude and frequency. Frequency can be thought of as the ‘pulse’ of the power system. When the pulse rate is within a healthy range, the power system runs smoothly. However, a pulse rate that is above that range may indicate a rise in blood pressure, and below could be a sign of blood pressure dropping. Neither of these scenarios is where you want to be. The same goes for the power system.
Named after Heinrich Rudolf Hertz
, the Hertz measurement (Hz) refers to the number of cycles per second. The 50 Hz power supply we receive at our homes in Australia alternates between positive and negative values, providing 50 cycles every second. In other parts of the world (e.g. North America) the nominal system frequency is 60 Hz. In an Alternating Current (AC) power system in its normal or healthy state, all the connected generators rotate in synchronism, and the speed of rotation of generators determines the frequency of electricity generated. In Australia, electricity is generated at a frequency that is very close to 50 Hz.
For one 100-second period, Figure 1 shows the measured normally-occurring variations of frequency at four well-separated locations in the National Electricity Market.
Figure 1. Measured variations of frequency at four well-separated locations in the synchronised part of the National Electricity Market.
The plots of frequency variations in Figure 1 show that:
- There are always small, random frequency variations around the target value of 50 Hz.
- The frequencies at the different locations in the interconnected system are almost the same.
The target frequency of the power system is set to around 50 Hz because that is the frequency that many of the large industrial energy users connected to the network require for their machines to operate. Gas turbines for example, do not perform very well when they are not connected to a power supply of 50 Hz or very close to 50 Hz. Therefore, the power system frequency can be considered the ‘pulse’ of the power system, and is one indicator of power system health.
Balancing generation and demand: power in versus power out
A healthy power system operating with a near-constant system frequency requires total generator electrical power output (to supply the loads and losses) to be equal to the mechanical power input to them. The difference in these two powers is provided by changes in the kinetic energy of all the synchronous generators in the system.
The kinetic energy of a generator represents the energy stored in the rotating mass of the generator (including any connected turbine), and is proportional to a property known as inertia. Therefore, release of kinetic energy as electrical power generated, is known as inertia response.
When the input and output powers are mismatched, a corresponding change of the speeds in all the generators in the connected system occurs, leading to a change in the system frequency. As our power consumption is always changing, so the total power demand through the system is constantly changing. This creates power mismatches and frequency variations like those shown in Figure 1.
If a large generator is disconnected from the system, a significant power mismatch is produced, and the system frequency reduces at a fixed rate. If there is no change to the mechanical input powers of the generators that remain connected, the frequency will continue to fall at that fixed rate until a system collapse occurs. More input power (or less power demand) is required to arrest the reducing frequency and stabilise the system.
Figure 2 shows the frequency in Adelaide on 14 March 2005, when different events produced power mismatches and substantial frequency variations. A system event caused sudden reductions of the mechanical powers to two large generators in South Australia. This produced a drop in system frequency and an increase in the power flow from Victoria. About two seconds after the initial disturbance, because of the increase in power transfer from Victoria to South Australia beyond the capacity of the VIC-SA interconnector, it was automatically opened, leaving South Australia electrically isolated from the rest of the National Electricity Market.
Figure 2: The measured variation of frequency in Adelaide following a sequence of events that left South Australia electrically isolated.
After the trip of the VIC-SA interconnector, the Adelaide frequency fell at a rate of about 1.3 Hz per second.
In this 2005 case, the fall in Adelaide frequency was subsequently arrested by actions that restored the power balance in South Australia.. Generators increased output power and some loads were disconnected.
In September last year, a very similar sequence of initial events (loss of SA generation plus tripping of the VIC-SA interconnector) occurred, but the consequence was quite different and the isolated South Australian system was blacked out. The main reason for the different outcome was that the kinetic energy released from the connected generators was much lower, due to the smaller number of synchronous machines connected to the South Australian system at the time, producing a much higher rate of frequency reduction and responses that were too slow to be effective.
The mechanisms for power balancing are commonly addressed by many inter-related terms such as; spinning reserve, Frequency Control Ancillary Services (FCAS), and system inertia response. Some aim to slow the rate of frequency reduction, while others aim to increase the ‘power in’ to stop the frequency drop and to bring it back to near 50 Hz. The mechanisms differ from each other, mainly in respect to the speed at which they can react, as shown in the following plots.
Frequency Controlled Ancillary Services agreements
Activating FCAS is the next step. The Australian Energy Market Operator (AEMO) has a number of FCAS agreements
in place, ready to go should the need arise. When the frequency is reducing, if the provider that has entered into an FCAS agreement increases their output they receive a financial incentive in recognition of providing an increased amount of power to support the system. The machines are programmed to increase the amount of (mechanical) ‘power in’ as the frequency goes down. The generators measure the frequency and know when they need to increase their power. They are always ‘checking the pulse’ of the system.
There is no human intervention during the event; it is all controlled using technology and everything is programmed. As soon as a generator enters into an agreement they activate a control kit. The FCAS contracts are managed by AEMO.
Some machines can provide power in about six seconds, some machines can do it in about 60 seconds, and some machines can do it in five minutes. So, the faster they can provide it the better, but it depends on the technology within the machines.
Fast Frequency Response to manage the energy system of the future
When we don’t have enough of those rotating machines across the system, for example when coal or gas fired power stations close, the amount of inertia available in the system reduces and it becomes difficult to keep the frequency constant, particularly just after a disturbance. With the retirement of several large generators scheduled to occur within the next 5 years, the number of machines that can increase their powers within the required 6 second window is also reducing.
One challenge is to find a way of providing something equivalent to stored kinetic energy into the system within a short period of time, specifically, referred to as Fast Frequency Response (FFR). As we transition to the energy system of the future, a number of possible solutions are being investigated, including the installation of synchronous condensers, and large-scale battery storage. Another possibility is for wind farms to make use of the kinetic energy stored in their rotating turbine blades, when the frequency falls. This process of providing ‘synthetic inertia’ requires the wind farm to reserve some of its output power capability, so that it is available when needed.