26 August 2013, Lagos – System collapses in the Nigerian power supply system have been reported in the newspapers. Up to thirteen have been reported this year alone, as at 1st of August. But what is a system collapse, what are the consequences, et cetera?
‘System collapse’ is the term used to describe the situation when all the power generating stations connected to the grid shut down at the same time or immediately one after the other, leaving the entire area supplied by the grid (the whole country, in the case of the Nigerian grid) in blackout. Generating plants may shut down as a result of human deliberate control action as part of normal operational scheduling, or automatically as a result of self-protective action.
Self-protective shutdown of generators may be initiated by protective relays installed on the generating plant reacting to any one or combination of potentially harmful operational development within the power station or individual generator. It may, on the other hand be as a result of happenings on the grid, external to the generating plant, but which have the potential to impose operational duties on the generator beyond its capability. One of such is when the system operating frequency falls outside the allowable deviation from the nominal level stipulated for the generating plant.
The system nominal frequency applicable to all generating plant and electrical appliances in Nigeria and many other countries is 50Hz (50 cycles per second). (The exceptions are in the Americas, north and south, where it is 60Hz). Thermal stations (i.e. those burning some form of fuel, especially the steam stations), due to their many pumps and motors controlling critical functions, are more sensitive to low frequency than hydro stations. But all of them, whether hydro or thermal, will eventually shut down when the frequency falls low enough, and then a system collapse results.
We may ask what causes the system frequency to fall. A load demand imposed on the grid greater than the combined capacity of all the connected generating stations will cause the system frequency to fall. With our low generating capacity (4000MW) in comparison to our country’s actual demand, it is obvious we will always be operating in the vicinity of low system frequency.
When the increase in load demand is due to consumers, the rate of fall in the system frequency is gradual and the speed governors of each generator can take some corrective action automatically by boosting the fuel input to the generator to increase its speed, and hence the frequency, provided the generator had not been fully loaded up to that point. This is similar to when, upon approaching a hill, the driver of a truck changes gear and presses down harder on the accelerator pedal of the truck.
Good system operating practice requires that one or more generators in the grid are deliberately only partially loaded so there is ample and readily available spare capacity, called “spinning reserve”, to absorb unplanned sudden load demand placed on the grid such as when a huge industrial consumer comes on line, or a generating plant is shut down by its own self-protective action thereby placing a heavy load demand on the remaining generating stations.
When these automatic actions, i.e. the combination of the speed governors and the spinning reserve, fail to arrest the frequency drop, the system operators at the National Control Centre who are monitoring the state of the power system and noticing the dropping system frequency, can intervene ‘manually’ and instruct one or more power stations to start and bring up additional generating plant.
They can in addition instruct a substation to disconnect some consumers to reduce the system load demand. Nowadays the system operators are not able to establish communication with stations quickly enough to arrest the impending collapse. This is because the SCADA (i.e. Supervisory Control and Data Acquisition) that enables the system operator monitor happenings at all grid stations, as well as the special “no fail” communications facilities built for instant linkup between the control centre and all stations when the grid was being developed in 1960s -1980s and which were still functioning up to the mid 1990s, are all now virtually broken down or not even there anymore. T
he voice communications have been largely replaced with GSM phones but their ‘network problems’ and ‘subscriber not available’ messages, and other interruptions obviously make the use of GSM completely unsuitable for this function.
Another scenario in which the system frequency falls is when the system overload is caused by a fault (i.e. a short-circuit) on any one of the many transmission lines that make up the grid. Since many thousands of kilometers of our transmission lines pass through heavy forests, the possibility of vegetation fouling the lines and causing faults is ever present, except the line trace is cleared of all vegetation up to 15metres on either side of the line for the whole length of the line, which is easier said than done, especially because of the heavy costs and logistics involved.
These line faults represent very heavy load demands that require every generating plant on the grid to contribute multiples of its capacity, which of course, is impossible. Here we have a “double jeopardy” situation. Apart from the overload placed on each generator, which is harmful on its own, the system frequency also drops very sharply and beyond allowable thresholds.
The system protective relays should normally remove the faulted line from service promptly before the generators’ self-protection relays react. But they can only do so, if they act in a coordinated manner, i.e. if they are properly calibrated and their settings coordinated with each other, and also provided the communications links between them are functioning. Again it is doubtful if this is still so nowadays. Protection engineers, specialists in their own right, whose function it is to calibrate and ensure proper coordination of these relays are an endangered species in the power industry nowadays.
Also, as has been mentioned earlier, most tele-protection communication channels in the power system are no more functioning as they used to. Line faults are therefore usually removed too slowly. The ‘last ditch’ provision to save the system from collapse in this situation then is the underfrequency emergency load-shedding arrangement. But it is not always very effective because the consumers on the feeders it removes from service may not have their total load anywhere close to the demand imposed on the system by the fault, and a system collapse still occurs.
Now why do blackouts due to system collapses last so long? The major problem encountered during restoration following a system collapse is the time required to restart the generating plants. The thermal stations pose the biggest challenge. They cannot remain spinning when they have no load connected to them and usually shut down when system collapses occur. And when they shut down, they eject their steam they have built up and depend upon to operate.
This is in order to avoid dangerous steam pressure buildup. It takes about six to ten hours to raise new steam and start operation again. And when they are restarted, they have to be loaded in small increments, allowing them to stabilise at each load level before taking on additional consumers. Sometimes the anxiety to load the generators quickly and shorten the system blackout may lead to taking on too much load at a time and cause a generator to shut down again and return the operators to the beginning all over again. All this while, the national control centre, NCC, is abuzz with activities, talking to and passing instructions to all the power stations and grid-connected substations across the country, coordinating the restoration efforts.
Even as simply as one has tried to describe it, the foregoing should have shown that maintaining the integrity of the grid consists of a series of many and painstaking activities, many of them highly specialised. It also requires the continuous availability of many facilities for the operators of the grid, as has been mentioned earlier, over and above simply building of power stations and substations in various parts of the country, and transmission lines to link them to each other. These facilities include but are not limited to the Supervisory Control and Data Acquisition, SCADA, and no-fail voice communications that must work even in the face of national power blackout.
And having made this point, one should also add that while it may not be possible to totally avoid system collapses, their frequency can definitely be drastically reduced. In fact in 1992 the country recorded zero system collapses. In the few years preceding that year (1992), through the determined efforts of the staff and the cooperation of Management (through the provision of funds), the coordination of the protective schemes in the grid was highly improved, most of the inconsistencies in the relay settings having been ‘debugged’; the SCADA having just been commissioned and put to use in the country’s power system for the first time, and the power system communications facilities being maintained in correct working order. Staff at that time were also very well trained to have been able to do the operation and maintenance work required at the highest standards. So if we really want to reduce or eliminate system collapses, government needs to look into all these.
By Engineer Foluseke A.Somolu, a formerly top NEPA official and Senior Special Assistant on Power Sector Reform to former Presidents Olusegun Obasanjo and Umaru Musa Yar’ Adua.
– This Day