Chapter 3Dynamic Demand Control
In this chapter the concept of dynamic demand control (DDC) will be introduced as the first potential DSM solution to the power system issues caused by increased deployment of renewable generation and electric vehicles. The basic control methodology will be demonstrated, and then experimental results from a laboratory scale DDC system with simulated wind generation and an EV charger. A simulation where DDC is used in a slightly more complex power system will also be shown, and conclusions made about the expected performance of DDC in an actual power system.
DDC is a method of DSM originally patented by Fred Schweppe in 1982 . In operation it works by simply adjusting the amount of load taken based on the measured system frequency. In simple terms, a power system can be considered as a rotating mass with various torques applied to it. Electricity generators apply positive torque so as to increase the frequency, and electrical loads apply negative torque which decreases the frequency. When these torques are equal, the frequency stays constant, and when there is an imbalance, the frequency will change. Normally a power system will have frequency responsive generators that constantly adjust power output in order to keep the frequency at the nominal value, which in New Zealand is 50Hz. As mentioned in Chapter 2, additional renewable generation can require additional frequency keeping as the generation output is often unpredictable.
Since system frequency is therefore a direct indication of the balance between supply and demand, adjusting the load in response to frequency should result in a stable operating point, provided that the combination between latency and system inertia allows it. One such example of a DDC scheme could be where a load draws zero power at 49Hz or below and full power at 51Hz or above as shown in Figure 3.1. As will be shown later, the stability of a DDC system is dependent on the system inertia. The lower the system inertia, the wider this frequency band has to be in order to maintain stability.
DDC Control Structure
Consider a stable power system, operating with some set base generation and base load, as well as varying generation, varying load and frequency responsive load. If the fixed items are ignored, this can be simplified in the control diagram shown in Figure 3.2, with symbol definitions as given in Table 3.1.
The additional low pass filter has a number of benefits. Firstly, if inductance machines are being used to generate electricity, there are some non-linear effects that must be considered. Sharp changes in the torque applied to the machine cause a change in the phase of the voltage waveform due to the armature reactance, and the measured frequency can therefore shift temporarily. If small changes in measured frequency cause a large shift in power consumed (and therefore torque) by controllable loads then the system could become unstable. In addition, if very simple frequency measurement circuitry is used, then noise on the voltage waveform could produce erroneous measurements with unknown results. By adding a low pass filter, the impact of these effects can be dominated by the filter time constant and the system can be kept stable.
The system performance is therefore dependent on the system inertia, filtering constant and available controllable load as given by (3.7). High gains and short time constants giving rapid response and high accuracy come at the expense of a low damping factor that is not acceptable. Thus the essence of control here is to have enough inertia in the local power system. Experimental validation has shown that using a digital low pass filter with a maximum delay of 0.16s gives an acceptable response for inertias of greater than 0.02 kG.m2/2-pole kW. Since inertia increases as the 4th power of the machine diameter times the length, inertias are more readily achieved with larger machines. This is one of the reasons why DDC is easier to implement in large systems with high inertia.
DDC Experimental Performance
A classical DDC controller has been tested under laboratory conditions and by computer simulation. In the laboratory a controlled AC drive in a torque limited mode generated a string of random torques changing each second to simulate a variable wind regime. The AC drive (variable speed induction motor) was connected to a 3 phase alternator generating at nominally 50Hz to form a motor generator (MG) set. Two of the phases were on resistive loads, and the third phase was passed to a DDC controlled load set up to emulate an electric vehicle battery charger. The VSD, loads and MG set are shown in Figure 3.3. Here the DDC controller adjusts the charging rate to keep the frequency at the nominal 50Hz.
The measured and computer simulated outputs are shown in Figure 3.4. An advantage of this experimental set-up is that the same random sequence can be used for all of the tests thereby allowing comparison of performance for different system parameters. In this experiment the operating frequency is allowed to vary and the system inertia is an important part of the damping factor for the controller.
In Figure 3.4, plot (a) shows the random torque signal, (b) shows the generator frequency, and plots (c) and (d) show the simulated and actual power consumption of the DDC load. As can be observed, the simulated output created with MATLAB Simulink® is a close fit to the measured data with the same average current and slightly less variation. This shows that the inertia figures for the experiment and the simulation are not quite identical and that the DDC controller behaved as expected to stabilise the system frequency. As previously mentioned, the stability of a DDC system is dependent on the system inertia. If DDC can be used to stabalise the operating frequency in a very small system, it will likely work even better in larger systems, especially as inertia scales exponentially with generator size.
Simulated Grid Performance
Further simulations were conducted in order to assess the performance of DDC in a grid environment. As mentioned in Chapter 2 there are a number of situations where a DSM system could assist in the operation of a power system. Of particular interest is the performance of DDC when integrating both renewable generation and electric vehicles. In the simulations here these were introduced both with and without DDC such that the results could be compared. The layout of the simplified power system simulated here is shown in Figure 3.5. Only 20 houses were simulated, each with a hot-water cylinder, electric vehicle charger, refrigerator and some base load.
Chapter 1 Introduction
1.2 Low Voltage Network
1.3 Electric Vehicles
1.4 Demand Side Management
1.5 Research Contribution
1.6 Thesis Outline
Chapter 2 Demand Side Management and the Electricity Grid
2.2 New Zealand’s Electricity Grid
2.3 Integrating Wind into the Power System
2.4 The Low Voltage Network
2.5 Transformer Overloading & EVs
2.6 The Future Smart Grid & Demand Side Management
2.7 Future DSM Options
2.8 DSM at the Transformer Level
Chapter 3 Dynamic Demand Control
3.2 DDC Control Structure
3.3 DDC Experimental Performance
3.4 Simulated Grid Performance
3.5 Performance of DDC
Chapter 4 Localised Demand Control
4.2 Controllable Loads
4.3 Benefits of LDC
4.4 LDC Control Structure
4.5 LDC Performance Requirements
4.6 LDC and DDC Comparison
4.7 Simulated LDC System
4.8 Load Scheduling Simulation
4.9 LDC Performance Discussion
4.10 Comparisons with other work in the literature
Chapter 5 LDC Laboratory System
5.2 LDC Experimental Performance
5.3 A combined DDC and LDC system
Chapter 6 Control Signal Transmission System
6.2 Star point injection system
6.3 Filter Design Criteria
6.4 Chosen Filter Implementation
6.5 MATLAB Design Verification
6.6 MATLAB Noise Rejection Simulation
6.7 Filter Performance in LDC System
6.8 System Performance Discussion
Chapter 7 Suggestions for Future Work
7.2 Practical Smart Loads
7.3 A Proposed Smart Appliance Interface
7.4 High Frequency IPT Power Supply
7.5 Electronic Transformers
Chapter 8 Thesis Outputs
8.1 Summary of Conclusions
8.2 Research Outputs
GET THE COMPLETE PROJECT
Localised Demand Control for Improved Grid Integration of Renewable Energy and Electric Vehicles