Why Standards are not the Solution

In this episode of Risk! Engineers Talk Governance, due diligence engineering experts Richard Robinson and Gaye Francis discuss why Standards are not the solution.  

They explore why following Standards can be problematic, outlining key issues: 

1.     Using Standards out of context;

2.     There’s no explanation for rationales;

3.     They don’t cover all things; and

4.     They are often lagging indicators.

They share a number of examples within SIL, AS1940 and AS61508 that demonstrate the issues if you do blindly follow the Standard. 

The episode concludes with their answer to an often asked question: If you don't design to Standards, how do Engineers design? 

The text they mention is Engineering Due Diligence (Concepts, Applications, Tools & Techniques) that can be purchased online at https://www.r2a.com.au/store/p/r2a-engineering-due-diligence-textbook  

For further information on Richard and Gaye’s consulting work with R2A, head to https://www.r2a.com.au.

Gaye is also founder of women’s safety workwear company Apto PPE if you’d like to check out the garments at https://www.aptoppe.com.au

Transcript

Megan (Producer) (00:01):

Welcome to Risk! Engineers Talk Governance. In this episode, due diligence engineers Richard Robinson and Gaye Francis discuss that Standards are not the solution.

(00:13):

We hope you enjoy their chat. If you do, please give us a rating and subscribe on your favourite podcast platform. If you'd like more information on R2A's work or have any feedback or topic ideas, please head to the website www.r2a.com.au.

(00:32):

And we do apologise for any background noise. We've done our best to remove it in the editing process. Thank you.

Gaye Francis (00:41):

Hi Richard. Welcome to another podcast session.

Richard Robinson (00:44):

Hello Gaye. It's good to be here.

Gaye Francis (00:46):

Just before we start, we've just noticed a bit of construction noise outside, so we apologise for that if it comes through in the background. We just had an observation last week that there was some noise coming through some of our podcasts. We are in an office space that's shared and sometimes we get some background noise from the corridor and outside. So apologies in advance for this particular session.

Richard Robinson (01:11):

For those of you who care about these things, we actually do try to minimise this because we're using a Rode podcaster too. And these are PodMics which are dynamic mics, which are obviously close field proximity devices to deliberately use to avoid background noise. But nevertheless, if there's enough noise out there, it will come through.

Gaye Francis (01:28):

So that was just a little bit of introduction, but today's session we're going to talk about Standards are not the solution. Now, we've talked about this probably a number of times in different podcasts throughout our seasons in different ways, but we're trying to bring some of this together because we're still finding that it's a big problem in industry. And some of the issues that we're asked to solve is because people rely purely on Standards.

Richard Robinson (01:55):

And we don't get it because I mean there's two fundamental problems here, well maybe three depending on how you look at it. But the first one is that people use Standards out of context, and that one I think is the one that mystifies us the most. And we'll give some examples of that. But I might just articulate those three. The first one's out of context. The second one is the Standard doesn't explain why that decision was made in the Standard. I mean, I keep coming across Standards to say the separation distance should be this. Well, was that based on tests or some conceptual design?

Gaye Francis (02:28):

Or a random number?

Richard Robinson (02:30):

Or random number? I mean, one of the ones I think I might've told you, I was coming back from a fire conference with Ted Ramsey from CSIRO, I forget why, but we're in the car together driving back and we were talking about standards and standard tests and things like that. And he was explaining this as one of the more delightful ones where the Americans devised this test for smoke development and basically they had this little chimney stack, which if you happen to measure with various Heinz cans on top of each other and you've got a small sample that you put on the bottom of this chimney stack, and then you put a sort of defined heat source underneath it, and you saw where this created certain amount of smoke coming out of the stack. She said there was one particular substance that burst in such violent flame that it belched out the front and didn't go out the stack at all. So we got a zero smoke rating even though it was clearly highly combustible, which is just absurd!

Gaye Francis (03:16):

It exploded, but no smoke! It's all right. It was just was an explosion, no smoke.

Richard Robinson (03:21):

Just another point. I've had my first experience with Covid in the last couple of weeks, so I've got a cup of hot water here, and if I suddenly start coughing or being a bit croaky, that's the reason why.

(03:33):

Yeah, so, I had three reasons and I forgotten what the third one is! <laughing> Oh, I know what it was. It was the fact that Standards don't cover all things anyway because we've had this experience a number of times. But going back to the first one.

Gaye Francis (03:48):

I'm going to add one more in there. I also think that they're lag indicators and there often five to 10 years behind what industry is actually doing, potentially.

Richard Robinson (03:59):

Well, it's like AS1940, they'd only updated it in 2017 based on Bunsfield. It took them, I don't know, 10 years. I think Bunsfield was 2005. And in 2017 is a fair amount of time to actually get your act together!

Gaye Francis (04:13):

And you would be hoping that some of the lessons learned from Bunsfield and the controls that you could put in place to prevent a Bunsfield would've been happening within that time before the Standard was updated.

Richard Robinson (04:23):

Yes. Well, there's the other point though. AS1940 is the standard, I used to spend an awful lot of time with working on. We used to use finite element models to try and work out the fire exposure and work out how much cooling or what had to be sprayed on tanks and tank farms and things like that. But one of the other options that the Standard suggests if you want to protect something is just put it behind in a chamber, sort of reinforced concrete walls so that any exposure fire would be prevented from exit causing troubles. But I've got to say the whole point of AS1940 as far as I've ever been able to tell is that it was located in an industrial area in tank farms so that if one of these things did happen, it was well away from any residences. To go and apply AS1940 adjacent to residential areas is not something that, as far as I know, AS1940 ever contemplated, but it doesn't say that anywhere. I guess that's one of the frustrations we have with standards. The context of the standard isn't actually properly articulated. And if you dunno the context, you dunno whether the Standard's relevant or not.

Gaye Francis (05:24):

And which they apply, isn't it? The situations in which they apply. Just makes it impossible. And I think the engineering environment that we are living in at the moment, those situations are changing.

Richard Robinson (05:40):

So fast!

Gaye Francis (05:41):

There's not a single environment that you apply all of these standards. There's all these nuances between the engineering solutions and the engineering challenges that we're seeing. And that environment and context seems to be changing.

Richard Robinson (05:55):

Well, one of the ones we were talking about, for various reasons we've got sort of sucked into high pressure gas mains. And one of the reasons why this topic has come back to the fore is because people keep building high-rise apartments next to roads that have high pressure gas mains in them. And obviously you can have a fireball. It's rare because somebody most likely a guy with a backhoe digger has to dig a hole and then punch into the high pressure pipe. Now it comes at supersonic speed, so it'll shriek. It's almost certainly going to find an emission source, so it's going to be really probably quite sad for the backhoe driver who digs the hole. But it also means you've going to go a very high, it's possible, it's not ever going to be likely because these things are well controlled and so forth.

Gaye Francis (06:37):

Credible, we use the word credible. It's credible that this could happen.

Richard Robinson (06:40):

Correct. And it does happen. I mean they did it in Sydney City a while back. They punched a hole. And you might recall we had a problem with when they were building the casino here, they managed to punch the offtake from the Sydney, at least the Dandenong to Melbourne high pressure gas main, which created some excitement. But one of the things, there's no Standard to it. It's the question is, well, when you're designing a residential building and you realise there can be a fireball out the front, what do you design the balcony and the windows to be? Because clearly if it's all glass and you happen to be on the balcony when this thing happens because you're curious about who's digging a hole in the road or something you are going to be in serious trouble. The heat is so great, so fast that it can kill you. I don't know if you remember those studies you did in on behalf of the, what was the Office of Gas Safety in Collins Street a long time ago ago where we used finite element models and working out the fatality rate for an ordinary commercial load building with glass windows.

(07:42):

Anyway, so one of the things we said was, well, you obviously don't want a combustible balcony, you don't want a transparent balcony. You've got to have a balcony that would actually act as a shadow so that if a firewall errupts, you can just drop to the ground, which is your natural tendency and have some shielding. And moreover, you don't want a fully glass window facing out there. You want to have a certain percentage that's blank wall so that you can crawl to the side and get to a safe location. And we have no knowledge and we've never seen any Standard which tells us what that ought to be. And our design solution was let's go with 50% glass and 50% non-combustible construction.

Gaye Francis (08:19):

So basically it gave somebody a chance in the rare event that there was this fire ball.

Richard Robinson (08:26):

But that leads on to the third point, which drives me crackers. We said that we don't know what the ratio should be. It's just an inspired guess on our part. So anybody can have an argument about that if they so wish. But when you look at a lot of these Standards, they don't tell you where these design criteria come from. They don't tell you why the separation distance should be 15 meters between tanks. Why is that? I've never seen it described anywhere. It's one of my frustrations with the SIL Standard in particular. Safety Integrity Levels. They have these numbers going from minus one, minus two, minus three, minus four for low demand and then it goes to minus five, minus six, minus seven, minus eight for high demand or continuous control. Where did these numbers come from? Is there a scientific breakpoint as to why this should be the case? And so far as I know, and we've been to lots of SIL conferences and talk about these things, no there's not. Some expert group just decided let's go with order of magnitude changes here.

Gaye Francis (09:20):

They all also seem to have been in place for a very long period of time. So you wonder what the context was going back to that first point in which they were thought up in the first place and have circumstances changed enough that they're still relevant at all?

Richard Robinson (09:34):

Well, remember we had that really peculiar thing we were doing with the transmission lines of Tasmania. We were trying to work out how transmission line heights were determined. Apparently it was all decided in New South Wales. And when we asked New South Wales to sort of explain themselves, well nobody could and we couldn't work out whether that's because they didn't know or because those who decided weren't with us anymore, which I suspect is probably the latter. But we don't know. And in the end, the only way we could work it out so far as we can tell it all comes back to Roman Horseman because every time we actually went and had a look at the design criteria, bridge heights so far as we can tell, have always been established by people gathering along on a horse and not bashing their head as they go under the bridge. And Roman Horsemen worked that stuff out a long time ago. And so that's just an historical height that was sitting there. And so if you look at a conductor that's going over a roadway and then you add the flashover distance that you'd get if a high voltage lightning strike hits it, which is 500 kilovolts and you add that flashover distance to that bridge height, you actually get the height that's in the Standard! <laughter> Now, whether that's the case, we don't know, and if anybody cares to look it up, you'll find it all described in our (Engineering Due Diligence) text, one of those more interesting things that popped up. But if you ask an electrical engineer about or transmission engineer, why are conductors at this height? The answer seems to be, well we''ve always done it that way.

Gaye Francis (10:58):

And the Standard says so.

Richard Robinson (10:59):

<laughs> Now obviously, I mean AS61508 makes the remark that if you've been doing it for long enough and it seems to work, that's actually a reason to keep a Standard. But it's worthwhile explaining and I think all standards should do it. Why they believe whatever it is they think is necessary and a good thing why it should be that way. I get very annoyed with fire standards and separation added between buildings and things like that because you sort of see it as one meter or three metres or five metres. Is that because somebody has thought it's a nice round number, it used to be three feet, so when you updated it, why wouldn't you make it a metre?

(11:39):

But is that a design reason? Does that explain what the load is? I mean, I think I've sort of explained to you that if you're building a firewall in Australia, they tend to just accept a four hour standard firewall, which is designed to resist a four hour standard fire. Now there is no such thing as a four hour standard fire! And if for example, you have a roll paper on end, you can't put it out, you have to assume burnout, which means the firewall has to be a massive independent, freestanding structure.

Gaye Francis (12:10):

It's going to burn for more than four hours.

Richard Robinson (12:12):

It's going to burn for more than four hours. So this idea that you have Standards that make sense and people haven't been testing the standards, and it's one of the frustrations we have because Engineers Australia keeps encouraging engineers to put their intellectual property into Standards. And I think I've talked about this one before, but none of the American Engineering Societies are dumb enough to do this. If you create the Society of Fire Protection Engineers handbook, it's your intellectual property and you sell it and you hang onto it, and they do give reasons for it. Remember we did that a long time ago we did that US fires that it's when you really found out you were allergic to chemicals, as a Chemical Engineer.

Gaye Francis (12:48):

I do.

Richard Robinson (12:50):

And we were lucky because one of our then young engineers about as young as you at the time was reading the Society of Fire Protection Handbook and realized that if you have electrostatic sparking in a high HDPE, high oxygen container, you can get ignition. The thing itself was meant to be always wet. It's one of the reasons why the HAZOP didn't work because they were actually commissioning it and it never occurred to 'em that it wouldn't be wet. So they never considered that configuration and HAZOPs have these weaknesses that we could go into, but that's actually one of the problems that we've had. Now it's just one of those extraordinary things. We were very lucky that we had a very bright engineer who was reading this book and why she was doing that, I still don't understand.

Gaye Francis (13:35):

Well, I guess the question people ask us, Richard, if you don't design to Standards, how do you design? If standards aren't the solution, which the engineering profession seem to be promoting, that that is the solution. Well, what do we do?

Richard Robinson (13:50):

You've got to figure it out from first principles. You look at the design options, you work out which design option's, right? Then you get the Standard out and check it to make sure you haven't done something daft.

Gaye Francis (13:59):

So it is, it's setting up what the context is. Where is this thing going to be designed? Where is it going to operate? What are the conditions?

Richard Robinson (14:07):

It requires you to think and thinking is hard.

Gaye Francis (14:10):

It is hard. It is hard.

Richard Robinson (14:12):

It is really hard. I don't know why people don't understand that thinking is hard. That's why you get the kudos for having theoretically gone to university. And if you have been thinking hard for eight or 12 hours, you are exhausted. At least that's been my experience.

Gaye Francis (14:29):

<laughs> That's another subject. So what we would say is have a look at your context, have a look at your problem that you're trying to solve, but also the environment in which it's going to live, the people it's exposing, because some of those vulnerabilities need to be thought about harder and there may be different solutions for those particular things.

Richard Robinson (14:51):

Yes. Well obviously we have some more experiences more recently with network design solutions and things like that. And it's particularly frustrating when you realise that the people who are responsible for this have not been thinking through what the further possible options are and not made it transparent, at any rate.

Gaye Francis (15:07):

I also think that Standards are lagging indicators. They're not always up with, they may be what represents current good practice, but that might be 5, 10, 15 years old.

Richard Robinson (15:19):

Well that's like AS1940 and Bunsfield. Bunsfield, I think it was 2005 and the Standard wasn't updated 2017 and the introduction says this is why we did it. To take into account Bunsfield. That's a long time.

Gaye Francis (15:31):

That's a long time. So you ought to remember that Standards are lagging indicators. It's the absolute minimum that you start with, but we would say that you actually go back and check against it rather than starting with the Standard and designing from there.

Richard Robinson (15:44):

Correct. And I think I'm going to be explaining this in the court in the very near future and we'll see how that goes.

Gaye Francis (15:51):

Which I'm sure will be the discussion of another podcast in coming weeks.

Richard Robinson (15:55):

Indeed.

Gaye Francis (15:56):

So thank you for joining us today. We hope you found it interesting. As you can see, it is one of our frustrations and I think a frustration around the engineering profession as well as a whole. So thanks for joining us and we hope to see you next time.

Richard Robinson (16:11):

Thanks Gaye.