At night, in a large open field, one can chance upon hundreds of individual blips of light from fireflies. As time progresses, it has been observed that these blips of light start synchronizing with each other, first in small clusters, then as larger groups until all fireflies are in synchronism and order appears to emerge from chaos. But what’s really going on? Is an explicit underlying communication channel or master signal required? Or can synchronization and harmonious existence be achieved just by each firefly adjusting its ‘discharge’ timing to align with its neighboring fireflies?
For the electric power system over many decades, frequency has been the global signal and the underlying communication channel by which synchronous machines aligned the position of their rotors and autonomously responded with their governor controls. This signal, available without additional infrastructure in a conventional power system of synchronous generators and loads, is implicitly linked with the quantity of power generated and the quantity of power consumed and provides a measure of the balance of the system. In fact, using electrical frequency as a communication signal has significantly shaped how we operate the bulk power system. While system operators also require the ability to periodically communicate with each generation source for the reliable and economic operation of the system, the faster and autonomous actions of resources use the system frequency as their communication channel.
As power systems move towards 100% inverters, the use of frequency as a communication signal can be questioned. The importance of maintaining electrical frequency is not being questioned, as there are thousands of electrical devices in the power system that depend on a steady frequency. However, frequency is not inherent to inverter-based generators and loads in the way that it is to synchronous resources. So as the importance of frequency as an indicator of load-generation imbalance reduces, can other means of achieving power sharing among sources be achieved? An introduction to this alternate form of control was discussed in a previous blog entitled “Ignoring electrical frequency in an all converter power system”.
In the firefly analogy provided above, it is possible for all fireflies to synchronize together without any global signal just by each firefly observing its neighbor. The instant a firefly’s neighbor emits light, the firefly would hasten to emit its own light. In this way, as each firefly tries to “catch up” with its neighbor, they end up becoming synchronized as shown in the figure below. In the figure, initially all fireflies are maintaining their own rhythm (top sub-figure). However, over time, they autonomously become synchronized (bottom sub-figure).
In the power system, a means of observing one’s neighbor is possible by using the angle of the voltage at the point of interconnection. Voltage angles, just like voltage magnitudes, are local signals in the power system. Based on the deviation of the angle from a reference, it is possible to share power among various sources. This principle is the basis for obtaining a distributed slack bus power flow solution and is fundamentally the mechanism that is used for controlling a synchronous machine’s power output.
The same result can be achieved by controlling the angle of the voltage phasor generated by the inverter. This angle, relative to its neighbor’s angle, could be used by the inverter to control the amount of power injected by the inverter source into the grid. However, how should this reference value be selected? In today’s grid, using frequency as a global communication signal with the entire system operating at the same frequency, the reference value for each source is the same. When the system operator wants to dispatch the generation resources to a different level, they just send a new value of MWs and a voltage magnitude.
However, when using angle as a means of sharing power, the reference set point in the controller of each source is unique and pertains only to that source. Would it then be possible for a system operator to re-dispatch these inverter resources as they do today? Or would the operator have to not only send a value of MWs and voltage magnitude, but also a voltage angle reference value? And even then, would it be the value of the angle at the point of interconnection (POI) of the plant or at the terminal of the individual inverters? If the angle reference is at the POI, then it would be the job of the plant controller to dispatch appropriate reference commands to each individual inverter.
In a recent ongoing research project at the Electric Power Research Institute, Inc., USA, the viability of using the values of local voltage angles to balance the bulk power system has been studied. The controller of each inverter resource continuously tracks its own angle as a reference. If the system is in a stable state, the controller updates the reference value of the angle to the present value. However, upon the occurrence of a disturbance, the change in the angle would cause the controller to latch onto the previously known reference value, and use the present value to obtain the amount of deviation which is then used to change the active power levels of the inverter resource. Once the system is back in steady state, the controller resumes tracking of its angle to set the reference value. In this manner, the system operator must again only provide a value of MW and voltage magnitude to each source.
Additionally, power balance across multiple areas can also be achieved using this control, very much like today’s secondary frequency response (or AGC). The difference is that the area control error is evaluated only based upon the deviation in tie line (inter-area) power flow, unlike in the present-day system where deviation in frequency is also a critical component of area control error. The deviation is measured every few seconds, and a power command is sent to each participating inverter resource a couple of seconds later. From a system operator’s perspective, for long-term and short-term power balance across the system, the number of control signals or references that need to be sent to each inverter resource remains the same as in today’s power system. By using this method, it is possible to share power among resources and operate the system at constant frequency.
Additionally, research results have shown that even if a small percentage of synchronous machines are present (assume base loaded machines), this kind of control is able to balance the bulk power system without adverse reactions on the synchronous machines. Further, as this control scheme strives to operate the system at constant frequency, it would be a seamless transition from this control scheme to a conventional frequency control scheme (using frequency droop). The need for transitioning between the control schemes could arise due to a large presence of synchronous machines in particular hours of the day. However, the transition back from a conventional frequency control scheme to this control scheme would have to be carried out carefully in order to prevent unwarranted transients in the system. The transition back would have to be carried out only when the conventional frequency control scheme is operating with the frequency near the nominal value.
However, this control scheme does have its drawbacks. Upon the occurrence of a fault or a network topology change, an active power-angle control loop is much more sensitive to the changes than is an active power-frequency control loop. It can also be said that by controlling angle, one is implicitly controlling frequency, and thus frequency is still being used as a communication mechanism. While this is certainly true, the aim of the research work is to identify ways in which a nearly 100% inverter system can coexist with a fully 100% inverter system, when electrical frequency would no longer have an inherent link to load-generation imbalance. This research work is still in its nascent stages and the road to an end solution is long and peppered with obstacles. The aim of the work is to prevent shutting the door on these possible alternate control schemes, as we continue to explore the possible capabilities from inverters.
Deepak Ramasubramanian
Electric Power Research Institute (EPRI)
Waqarul Mulk says
Hi
I am installing an ongrid inverter and I want no import in bulk and zero export,but just want to take a small amount of reference source from grid to run my load. Is there any way,so that I can proceed.
Thanks