I have known for a long time that the “bright-line” criteria were anything but bright, and that there were going to be a large number of questions when it came down to actually applying them. Well, guess what? People are applying them now, and there are a large number of questions. This is the first in what will probably be a number of posts (not all together, of course) discussing bright-line criteria problems.
I used to think there would be a guidance developed for using the BLC (beyond what is in the Guidance and Technical Basis section of CIP-002-5.1, which is good but simply limited), but I certainly don’t see that happening now. I imagine these questions will be answered the way most other CIP v5 questions will be answered: by each individual region (or even each auditor within each region), with great variability across regions. That’s why I call them the Not-so-Bright-Line Criteria.
The reason why it’s inevitable there will be many questions on the BLC is that the electric power industry is tremendously diverse. I am always struck by the differences between facilities (e.g. one generating station vs. another, one substation vs. another), between the NERC entities themselves (obviously, coops vs. munis vs. IOUs, but even within those three categories there is huge variability in what they do, etc. That’s why NERC has all those registrations, and why almost no two entities have the same registrations), and between the environments in which the entities operate (entities in PJM have an entirely different set of constraints than those under CAISO or SPP, for example). There is simply no set of criteria like the BLC that could cover this diversity.
I noticed early on (when entities started considering CIP v4) that just about everyone I talked to would say, “We think we understand where we fit in the bright-line criteria, except for….” And here they would rattle off a long story about how they have one substation or generating plant that might look like it meets one of the criteria but in reality it doesn’t because of some complicated reason that applies just in their region, etc. I figure that, if every NERC entity has at least one “except for” situation, and there are maybe 5-800 entities subject to CIP v5, there are a whole bunch of these issues out there, waiting for me to write about. So I expect to be done with this series of posts in about 2020, and then only because CIP v7 or v8 will be in force, not because all of the questions will have been answered.
Now on to the problem….
There are several strands to this problem, but the main one is that there is no NERC definition of “substation”. Of course, the SDT knew this when they drafted CIP v5, and for the most part the criteria in Attachment 1 of CIP-002-5.1 are free of any dependence on the word “substation”. However, that is not the case for criterion 2.5.
In that criterion, the Facilities that are Medium are those between 200 and 500kV. As everybody knows by now, there is a formula to weight the different lines in the substation, and the sum of those weights has to exceed 3,000 in order for these Facilities (i.e. the 200-500kV ones) to be Medium. If that sum is less than 3,000, then all of those Facilities, and the BES Cyber Systems associated with them, are Low impact.
Now suppose you have a substation with two control rooms, one that controls two 345kV lines and another that controls one 245kV line (whether this is likely to happen or not, I have no idea. I’m not a power engineer, and I don’t even play one on TV – assuming there are any TV shows about power engineers). The two 345kV lines are worth 2600 points in the formula, while the one 245kV line is worth 700 points. Together the substation has 3300 points, meaning that all the BES Cyber Systems associated with these three lines will be Medium impact[i].
But let’s say the two control rooms (and their respective lines) are in two separate substations. The point total for the first substation will be 2600, while the second will be 700. In this case, all the BES Cyber Systems in both substations will be Low impact, since they no longer meet criterion 2.5, and don’t meet any of the other Medium criteria.
You probably know where I’m going on this, but let’s go back to the original case where both control rooms are in the same substation. Now let’s start moving the two control rooms apart (since this is a thought experiment, we can do this in the blink of an eye. It would take somewhat longer than that – he says with a grin – to do this in real life).
Let’s move the second control room 200 yards away from the first. Are they still the same substation? If we think so, let’s move them half a mile apart. Are they still the same substation? How about if they’re a mile apart? Or is it the fence that makes the difference? As long as there are two fences between the two control rooms, could they be located just twenty feet from each other and be in separate substations?
Of course, there’s no answer to these questions, since there’s no definition of “substation”. But I predict this will become an issue as entities focus more on their compliance costs for substations subject to criterion 2.5.
Now let’s put another turn on the screw. For the remaining discussion in this post, I wish to acknowledge the assistance of a large transmission entity that brought this issue up to me and helped me (try to) understand what is involved.
Suppose there’s a “transfer-trip” relay associated with the 245kV line; this relay can trip one or more of the 345kV lines in certain circumstances. In the scenario where both control rooms are indisputably in the same substation, this doesn’t change anything; the BCS in both control rooms will be Medium impact.
But let’s now throw another 245kV line into the first control room, giving that control room by itself 3300 points under criterion 2.5. And let’s move the second control room – still containing the transfer-trip relay – 500 miles away, with 14 fences and a moat with alligators between it and the first control room; the transfer-trip relay in the second control room still can trip the relay associated with the 345kV line in the first control room (I realize this is an extremely unlikely situation in real life, and probably violates the laws of physics anyway. But stay with me). I don’t think anyone will dispute that the second control room is in a separate substation. Since that substation only has 700 points under criterion 2.5, and since it doesn’t meet any of the other Medium criteria, it will be a Low substation (of course, the substation with the first control room now meets criterion 2.5, so the lines in the first control room are still Medium Facilities).
However, someone who hasn’t read my blog for the past three months might point out that Attachment 1 Section 2 of CIP-002-5.1 says that BES Cyber Systems “associated with” any assets or Facilities that meet one of the Section 2 criteria are Medium impact. Since the transfer-trip relay in the second control room is definitely associated with one of the Facilities (a 345kV line) in the first control room, it will be Medium impact, right?
Of course, the answer to this question is “no”, given NERC’s recent “ruling” on such situations, described in this recent post; the transfer-trip relay in the second control room will be Low impact. So now the question of what is a substation again rears its head: as we gradually move the two control centers closer to each other, at what point do the separate substations become one, meaning the transfer-trip relay will go back to being a Medium impact?
Again, I don’t have the answer to this question (that’s why I love being a blogger. I can point out all sorts of thorny issues and make people at NERC and the regions squirm, without having any responsibility at all to be constructive and actually solve the problem), but I hope there will be one answer given at some point (presumably from NERC) rather than eight or more answers (one for each region, and perhaps more for auditors within the same region who have differing opinions on this).
Now I turn the screw again.[ii] Let’s suppose the substation in question isn’t a criterion 2.5 one at all, but a criterion 2.4. It has a 500+kV line (we’ll say 765kV in this case) controlled by one control room, and a 345kV line controlled by the second control room. Once again, there is a transfer-trip relay in the second control room that can trip the 765kV line.[iii]
In this case, the situation is different from the criterion 2.5 case. Let’s look again at what happens if both control rooms are at the same substation. Criterion 2.4 says that “Transmission Facilities operated at 500 kV or higher” are Medium impact. This means the 765kV line is a Medium, but the 345kV line is Low (since the substation itself doesn’t figure in criterion 2.4, as it does in 2.5). However, since the transfer-trip relay is associated with the Medium impact (765kV) line, it will also be a Medium BCS.
We once again separate the two control rooms by 500 miles and alligators, so there are now two substations. The second control room will without doubt be a Low impact, but since the transfer-trip relay remains associated with the 765kV line, it will be Medium impact, right? Before you bring up NERC’s recent “ruling” referred to above, remember: that ruling only applies to criterion 2.5. We’re now dealing with 2.4.
But if you follow my blog closely, you’ll know that the August “ruling” – by Steve Noess of NERC - that I referred to wasn’t the first “ruling” on this issue. The first was in June by Tobias Whitney of NERC, as described in this post. Tobias reached the same conclusion as Steve, namely that a “far-end” or transfer-trip relay located at a Low substation but associated with a Medium line was Low impact, even though it is clearly “associated with” a Medium Facility.
However, the reason Tobias gave was quite different from Steve’s. Steve’s was based strictly on the wording of criterion 2.5 (and followed very closely the reasoning an Interested Party had provided me in a post I did the day before Tobias made his pronouncement). On the other hand, Tobias’ reasoning was much more general: he said[iv] something to the effect that “Physical location IS a determinant factor for impact classification.” Note there is nothing specific to criterion 2.5 in that dictum.
Let’s apply Tobias’ reasoning to the criterion 2.4 case. Given that the transfer-trip relay in the second control room is clearly in a separate substation (and it’s 500 miles away from the first substation), I’d say that if physical location is ever going to be a “determinant factor” for classification, it should be here. I say the transfer-trip relay has to be a Low impact one. So Steve’s and Tobias’ reasoning lead to different results in this case.[v] Which person should you believe? I would say Steve, since his reasoning is written down (in an email), while Tobias' reasoning isn't (his is in a PowerPoint, but that was never officially released after the meeting, which is known as having your cake and eating it, too). So the transfer-trip relay, which is located in a Low substation, will still be Medium impact - because the criterion we're dealing with now is 2.4, not 2.5.
Once again, let’s start moving the second control room back toward the first. Whenever it becomes part of the original substation, then there will be no more question that the transfer-trip relay is Medium impact (under either Steve's or Tobias' reasoning), and peace will be restored at NERC. Of course, the whole question is when the two substations do become one, which is the same as my original question: What is a substation? That question remains as unanswered as it's always been.
So we’re back where we started, but I hope this exercise has at least cleared up what the issues are. It certainly has for me, and I hope it will in some way lead to the two issues in this post (the meaning of "substation" and the Steve/Tobias difference) being addressed in some definitive or semi-definitive way by NERC. This was fun, and I would like to hear of any similar issues you may have discovered in the bright-line criteria. Contact me at firstname.lastname@example.org.
Sept. 13: I've just posted a sequel to this post.
The views and opinions expressed here are my own and don’t necessarily represent the views or opinions of Honeywell.
[i] Remember that criterion 2.5 says that “Transmission Facilities” operating between 200 and 500kV are Mediums, if they are associated with a substation that meets the 3000-point threshold. Since all three of these lines fall in that range, all the BCS associated with them will be Mediums. Of course, there’s an exception to this (Alrich’s Law states that no statement can be made about a NERC regulation that doesn’t have at least one exception. So far, I have found no exceptions to this law). That exception is the case of the “far-end” relays discussed in this post.
[ii] Of course, when I speak of turning the screw, I’m thinking of Henry James’ wonderful ghost story, “The Turn of the Screw”.
[iii] This is the scenario – perhaps mangled in my retelling – that the entity I referred to above brought to my attention.
[iv] At a NERC CIPC meeting in Orlando.
[v] My guess is that Tobias’ reasoning, since it could conceivably apply to other criteria as well, will lead to further contradictions with Steve’s reasoning. That is, unless someone at NERC states whose reasoning is the one to follow in this case.