September 07, 2004
Electric Power Grid Vulnerabilities Explored
Reka Albert, an assistant professor of physics at Penn State, has led a team examining the national electrical grid in the United States to look for vulnerabilities and her team has found that failures in a fairly small portion of the network can lead to a major disruption. (same article here)
"Our analysis indicates that major disruption can result from loss of as few as two percent of the grid's substations," says Albert, whose research team includes Istvan Albert, research associate in the Bioinformatics Consulting Center at Penn State, and Gary L Nakarado at the National Renewable Energy Laboratory. One implication of the research is that identification of strategic points in the grid system can enhance defense against interruptions, whether by equipment failure, natural disasters, or human activity. Major blackouts caused by failures in the grid, such as the one that affected the northeastern part of the country during the summer of 2003, incur tremendous economic, public-health, and security risks.
The study, titled "Structural Vulnerability of the North American Power Grid," was published in a recent issue of the journal Physical Review E. The researchers constructed a model of the entire transmission grid with over 14,000 "nodes," including generators, transmission substations, and distribution substations, and over 19,000 "edges," corresponding to the high-voltage transmission lines that carry power between the nodes. They measured the importance of each substation node based on its "load," or the number of shortest paths between other nodes that pass through it. "While 40 percent of the nodes had a load below one thousand, the analysis identified 1 percent of the nodes--approximately 140--that have a load higher than one million," Albert says.
However, the grid quickly becomes disconnected when the high-load transmission substations are selectively removed from the system--if the nodes that have the highest load are removed first, followed progressively by the nodes with successively lower loads. According to the model, a loss of only 4 percent of the 10,287 transmission substations results in a 60 percent loss of connectivity. During a cascading failure, in which the high-load substations fail in sequence, the model shows that the loss of only 2 percent of the nodes causes a catastrophic failure of the entire system.
Whether regulation can be changed to allow market forces to bring about changes to make the electric grid less prone to massive failure is a question that Lynne Kiesling addresses on her blog. My guess is that the costs from failures hasn't gotten large enough to overcome the status quo bias problem that Lynne has discussed recently. Also see her post Network Reliability As A Public Good And What To Do About It.
Advances in technologies such fuel cells, cheaper batteries, and Vehicle to Grid will likely lead to much more capacity for local generation of electric power in the longer run. So grid reliability will become relatively less important than it is today.
This is not news, by any stretch. The Feds have known about this sort of issue since at least the late 1980s.
While the authors above make a point about failure of a couple of percent of stations pushing the system into chaos, there are some points where knocking out a couple of large transformers could drop the entire system for *months*, or at least put it into an unstable mode which would induce widespread blackouts during high-load periods.
What's wrong with the electric grid? - The Industrial Physicist:
Snippets, though the entire article merits study especially by economists and legislative analysts:
..."Experts widely agree that such failures of the power-transmission system are a nearly unavoidable product of a collision between the physics of the system and the economic rules that now regulate it. To avoid future incidents, the nation must either physically transform the system to accommodate the new rules, or change the rules to better mesh with the power grid’s physical behavior....
..."In the four years between the issuance of Order 888 and its full implementation, engineers began to warn that the new rules ignored the physics of the grid. "The new policies do not recognize the single-machine characteristics of the electric-power network," Casazza wrote in 1998. "The new rule balkanized control over the single machine," he explains. "It is like having every player in an orchestra use their own tunes....
..."The net result of the new rules was to more tightly couple the system physically and stress it closer to capacity, and at the same time, make control more diffuse and less coordinated—a prescription, engineers warned, for blackouts.
"In March 2000, the warnings began to come true. Within a month of the Supreme Court decision implementing Order 888, electricity trading skyrocketed, as did stresses on the grid (Figure 3)...."
---- Lesson available for learning here?
For those who don't read the original, a bit more in the way of excerpts from that IEEE article
----> on the physical results of adding more generating capacity to the grid we have now:
'In the view of Casazza and many other experts, the key error in the new rules was to view electricity as a commodity rather than as an essential service. Commodities can be shipped from point A through line B to point C, but power shifts affect the entire singlemachine system. As a result, increased longdistance trading of electric power would create dangerous levels of congestion on transmission lines where controllers did not expect them and could not deal with them.
"The problems would be compounded, engineers warned, as independent power producers added new generating units at essentially random locations determined by low labor costs, lax local regulations, or tax incentives. If generators were added far from the main consuming areas, the total quantity of power flows would rapidly increase, overloading transmission lines. “ The system was never designed to handle long-distance wheeling,” notes Loren Toole, a transmission-system analyst at Los Alamos National Laboratory.
------> On the specific physical requirement of the system that you might call a "public good":
"Finally, the separation into generation and transmission companies resulted in an inadequate amount of reactive power, which is current 90 deg out of phase with the voltage. Reactive power is needed to maintain voltage, and longer-distance transmission increases the need for it. However, only generating companies can produce reactive power, and with the new rules, they do not benefit from it. In fact, reactive-power production reduces the amount of deliverable power produced. So transmission companies, under the new rules, cannot require generating companies to produce enough reactive power to stabilize voltages and increase system stability.
"The net result of the new rules was to more tightly couple the system physically and stress it closer to capacity, and at the same time, make control more diffuse and less coordinated—a prescription, engineers warned, for blackouts."
---> Much more in the original article, which I commend to your attention.
---> Looked at in the light of the known physical facts about how electricity works, deregulation appears to have been akin to defining pi=3.0 to simplify geometry. On paper, it looked okay.
The grid is is more stable then most people think it goes in and out a lot because of weather patterns all the time. It is quite easy to route around problems with automatic switching. The terrorists would have to know precice locations to strike based not only on loading conditions (time of year), but also based on the switching patterns available to operators in the control facilities of transmission companies. Also 0.04 * 10287 is roughly 411 substations. Assuming a small 3 man team for each sub that is still 1200 people. That's a lot of people and coordination. However, the biggest problem resulting from such an attack would probably be the time it takes to get a new power transformer let alone 400. This seems like an unlikely possibility; especially if you consider that they want to kill Americans not just inconvenience them.
What patent nonsense. I am stunned that these so called educated individuals would misapply network graph theory as they did. Electric power systems are NOT pipelines or road networks--graph theory has limited applicability to an electric power system. The tools of use should be load flow and (possibly) stability software. I imagine that these people have never heard of remedial action schemes or special protection schemes either. I believe they are probably quite ignorant of the planning or operating practices of a typical large utility. The electric power system in the United States is not third world, nor is it particularly vulnerable. In case you have not noticed Mother Nature beats up these systems on a regular basis and they seem to perform well enough.
The post about the level of effort to really bust up the power system is quite relevant--last I looked (out my office window) a substation is a BIG place, three people per station might be a little light to get the job done. Any attack like this would have to coordinated and timed perfectly as well--we would notice the lights going out. Attacks of this nature are things that only governments can pull off--we have a name for this, military invasions. After staring down the Soviet Union for 70 plus years I don't think some terrorists are going to be much of a problem to defeat.
Clearly the stabilization of voltage resides with power grid regulation. CLEAN power regulation.
the person who is saying it would take more than a person or two to disable a substation is, in my opinion dead wrong. A rifle is the tool needed and ONE person can cause serious damage and doesn't even need to be close by. Substations need to be walled in to eliminate the possibility of causing massive damage and massive outages with bullets. Cheap and effective solution to a major potential problem.