Even though we are satisfied that our own systems are either Y2K compliant, or that we have a satisfactory work around, we are still looking for all of those other "unexpected" items that might spring out to bite us. Today I have been busy working on two of these, and its possible that others in the electrical industry may find these of some interest.

The first one concerns a major scientific and research company (government owned) who have discovered that the program they use to supply us with vital meteorological and hydrological data may not be compliant. I emphasise that the program "may not be compliant" not that it "is not compliant". Unfortunately the programmer responsible died recently, and left very little documentation on the data transfer program. It appears to work in simulations, but can not be adequately tested for real. The solution is a new program that they are currently testing, that will supply us with the required data in the correct format. It does mean a small program change in our systems which we have already locked, and for which we will have to convince our Y2K systems manager that it is required. If we do not change, we will not lose any generation capability, but we may lose a small amount of efficiency, and we may also lose our weather forecasting ability. However the existing program may also be OK, and we lose nothing.

The second issue is much more important, and although it doesn't relate to the initial cause of any problems, it may affect the restoration process following a grid failure. Our group of hydro stations has a black start contract to reliven the grid in the event of a total blackout. The company responsible for the national grid sent us a restoration plan to follow which initially restored power to a major smelter, then restored other (domestic) comsumers later. On first reading through the plan it appeared OK, untill I realised the type of load that a smelter takes. It is a purely resistive load, with large incremental steps. Although our generators are quite large enough to supply the total demand, there appeared to have been no calculations made on the frequency and voltage swings that would occur by livening such a large consumer first. In fact it would very unlikely that we could keep our generators connected long enough to establish a supply. As a result it now appears that some areas of domestic consumers will be connected first in order to supply sufficient electrical and mechanical inertia to the system prior to connecting the smelter.

I'm not sure what black start capability or facilities may exist in USA, but in light of the calculations that I've had to carry out today, it may be prudent to enquire what the black start procedures are in your areas.

Malcolm

-- Anonymous, October 27, 1999

Answers

Malcolm: Black Start capability, and established plans & methodologies vary quite a bit throughout the U.S. California and some surrounding states have had guidelines & procedures in place for years. These procedures were reviewed & applied in training sessions at various companies by operational personnel and SOME actual testing took place. It was during some of this testing and scenario playing that many problems were identified, i.e., HOW MUCH load could an isolated generator pick up at ONE TIME....or how much power would be left to pick up load with a generator at the end of a 125 mile long line, etc....WHAT load was picked up...WHEN...WHERE. But there are many regions that only gave a superficial scrutiny to system restoration ("Oh we'll just black start that hydro unit and route it over to that plant & start it and START picking up the system..."). Until very recently there were some regions of the country that had NO system restoration plans! Just as 99% of the population (including many in the industry) have no idea how complicated the power system is, even more have no concept on how precarious & complicated restoration can be. As an example, should the Western U.S. shut down completely, it could be between 24 hours (at best) to days before all load is restored and this would be assuming all goes well and there are no MAJOR problems. Another issue, or fly in the ointment, can be deregulation. "Black Start Capability" was monitored & controlled under the old methods by COMMAND & CONTROL of various utilities. Under deregulation, this "ability" has been segregated and made into a "market "commodity". That all should be very interesting when the operators call & "order" a unit on line to carry "isolated" load, etc. Also, people might not know that getting consumer load back on is NOT (or should not be) the highest priority. A power system needs to restart many plants...return transmission lines to service...establish station Light & power to substations and get outside power sources to certain critical plants...BEFORE picking up customer load. And Malcolm, a fluctuating load like a Smelter would be one of the LAST loads I would want up...especially on a weakly established system. Good Luck!

Quenton

-- Anonymous, October 27, 1999

Quenton - In the case of a smelter or steel mill, being shutdown for more than a brief period of time is simply not an option. Molten aluminum and steel tend to solidify and really gum up the works quickly. If I'm not mistaken (and Malcolm can certainly correct me) a smelter in NZ was hit with a protracted outage a few years back and it cost millions to clean it up and get it back online. (Note: this was not due to failure of electricity in the plant, but plant process control).

While these types of loads may not be critical to public health and safety, in terms of economic costs both to the company and the surrounding community, the loss of power or shutdown of these plants can indeed be devastating to a community.

We've addressed prioritization in any power restoration situation before in this forum (you can search the archives) and it goes something like: Public safety (fire, police, hospitals, emergency services), high priority industrial and residential, then remaining interruptible / low priority commercial and industrial loads.

-- Anonymous, October 27, 1999

Sounds like both Malcolm and Quenton agree about this. So, a triage debate is about to erupt. Good. I agree that what Rick says is correct. Saving the industrial base is important. But, if trying to save that base actually creates another cascading default situation, then it looks like the plant will have to die. We aren't talking about having a lot of time to play with here, are we? Those plants will go stiff within an hour or so, right? Gonna be some hard decisions made here. Hope the folks who are in command at that time know what triage is, and how to save the system, even at the cost of some peripheral users. I don't know if this is similar, but about a month ago our lights started to flicker all over the house. I plugged a voltmeter in and watched it jumping back and forth between 122 and 126 volts, every second or two. This went on for about 5 minutes. Never saw that happen before. Could this have been a load problem on the system?

-- Anonymous, October 27, 1999

Rick,

You are absolutely correct about the smelter here in New Zealand, and it is the same one that is the issue for a black start. Its previous failure was due to the smelters process computer not recognising Feb 29th during a leap year, and shutting down all pot lines. Fortunately that experience has confirmed that if the power supply fails, we have 90 minutes to restore it before the aluminum solidifies. This particular smelter is New Zealand's largest single user of electricity, and is also the largest export earner in the South Island. Hence the priority to re-establish supply as quickly as possible.

The original restoration plan was: Open all line and transformer circuit breakers at all sites in the affected area. Black start one 120 MVA generator. Close in one local service transformer (about 4 MW). Liven one 220 KV line to our smaller hydro station. Start a second 120 MVA generator. Liven one line to the smelter (150 miles away). Liven the smelters local load (22 MW). Ramp up minimum supply to a potlines running gear (45 MW). Start a second pot line (50 MW).

It is this step that would probably cause the system to go black again.

The modified plan that we are pushing for is to liven up a small city near the smelter before supplying the smelter itself. This would add about 20 minutes to the restoration time, and would allow us to connect 750 MVA of generation before livening the smelter.

Hopefully nothing will happen and all our planning will not be needed.

Malcolm.

-- Anonymous, October 28, 1999

Rick, GP & all: First, I can't seem to get paragraph structure to work on my responses...must be something I'm doing wrong...I do write better than they look. Rick, your prioritization references is mostly at the DISTRIBUTION level....you MUST prioritize at the bulk power level first on a total system restoratione effort. In that capacity, the ONLY way you can restart your system is to use available startup ower to start the power plants, establish the transmission system & strength, power for computer status & monitoring (allows the operator to "see") and only then start picking up major load areas....carefully. This process is very involved, with many details, not noted here.

I was an electric utility representative at a DOE initiated workshop a few years ago where the major thrust was restoration & priorities after a complete shutdown. Representives from the petroleum insdustry, Aerospace, National Defense, Communications, Emergency entities, etc., attended. They ALL believed they were more important than the other and needed power first. We electric folks explained that power system hardware, generator limitations and line configurations would NOT allow such segregation on a system restart. The petroleum industry actually believed startup power should be bid on...(same problems as the smelter)...that started a ruckus!

To amplify another point of Malcolm's. Any generator that is started, routed to another plant (or any load for that matter) must be VERY careful in picking up load. Heck, an isolated 175 MW generator can only pick up load in 10 to 15 MW increments..or it will TRIP..and you will start all over again.

Another point on prioritization of load. Public Utility Commissions, and similar bodies, have established such pecking orders in years gone by (emergency entities, hospitals, police, etc)...but they were for distibution focus...to establlish a preclusion for manual load shedding, avoidance of tripping mechanisms, etc. They were NOT established for power system restart priorities!!!

Power system blackouts can be complicated..and restoration very time consuming & involved..no matter what we might desire or believe is possible.

Quenton

-- Anonymous, October 28, 1999

Q - you are right, the focus of my previous comments were on distribution restoration, not transmission, because I was responding to Malcolm's observations regarding restart of the smelter (which would be a distribution system load). I freely admit that I'm not a T&D "expert" - but it would seem to me that transmission and distribution systems restoration following a total regional blackout are linked at the hip.

I agree that prioritization of distribution system restoration is a tricky topic at best. With the exception of large commercial loads that have their own substations, it's not easy (or possible, in a lot of cases) to be as granular in restoration as restoring power to a fire station in a neighborhood but not the entire neighborhood itself.

-- Anonymous, October 28, 1999