Saturday 28 November 2020

"Energy is Big and Sexy" - BBC. Well, Yes

As you will imagine I have been idly watching the Beeb's Powering Britain, not in the hope of learning much, but rather to see how prime time TV covers - or dodges - some meaty issues.  Needless to say, for the most part the four episodes have been lavishly-photographed tourist guides, our breathless reporter always being gushingly overwhelmed by the scale and sheer sexiness of the whole thing.  Big Engineering always has that effect - if you steer people away from dirty, leaky old kit and fix their gaze firmly on the shiny new stuff.  Helicopter rides are generally quite fun, too (on a day when the weather's ok ...)

All in all, a massive PR opportunity for the firms involved (SSE, Drax, Spirit and EDF) which naturally they've seized with both hands and immense gratitude.  Can't have done the whole industry any harm, either - give folks an idea of the scale of what it all means; romance of engineering & commerce, etc.  Indeed, if a similar series had been run on commercial TV or a newspaper, you'd assume it was paid-for advertorial stuff.   

And not a Green in sight!   Dear me no - we're all quite green enough without letting Swampy or Greta come on with their whingeing and wimpering.

Controversy has barely been acknowledged: just the once, really, over Drax.  The SSE episode was about the world's biggest offshore windfarm and its onshore receiving station - including the merest hint (which probably passed unnoticed to most) about what's fast becoming a cause célèbre, the culpably chaotic business of digging big cable corridors, generally through highly sensitive coastal geography & habitats etc, with no obvious sign of planning / coordination on the part of National Grid.  Otherwise, it was just 100% jaw-dropping Big Kit on display in breathtaking marine vistas.  And no mention of what happens when there's no wind? ...

The Spirit episode featured their huge Morecambe Bay gasfield and its onshore gas processing plant.  (If you haven't heard of Spirit, it is a Centrica spin-off, one of the many new 'end-of-field-life' specialist O&G producers who manage upstream assets when development risk is long since past and the original developers - in this case British Gas - have better uses for their capital.)   Big offshore installations are pretty mind-blowing, so no shortage of gawping to be done here.  What controversy might have been expected?  Well, there are some people for whom even mention of fossil fuels in any other sentence than "we are closing this thing down as fast as humanly possible, ideally tomorrow morning".  But the Spirit PR team had put clever words into the mouth of the plant manager, who simply said that we'll need gas for a few more years on the path to Net Zero Carbon (tacitly answering part of the question left by the SSE programme), and that they were there to do their bit.  The Beeb felt no need to qualify that with any sort of counterpoint voice-over.  

The last episode was on EDF's nukes at Heysham, and the nuclear fuel plant at Springfields down the road.  They didn't explicitly use it to answer another part of the unasked SSE / wind question.  Obviously, they could have filled the entire slot with nothing but controversy (see, for example, the Public Accounts Committee report published today); so they ran with more-or-less none whatsoever.  Fair play, it had to be that way, really: though arguably they might have mentioned the cracks in one of the boilers and some of the fuel bricks ...  So it ended up quite pedestrian to my view.  I was, however, entranced to hear engineers talking in "thousandths of an inch" (it was the same in the Drax episode) which jars a little.  Then again, I imagine they run the plant on Windows 98 or some such.

So what about Drax, then?  Yet again, they didn't use it to answer the SSE / wind question (- it actually contributes to both parts, in fact); but, yes, they really couldn't - and they didn't - fail to mention that burning trees to generate electricity is controversial.  Which it bloody is - an outright scandal, in fact, compounded by the risible official "green" carbon-accounting convention which allows Drax to ignore CO2 emitted at the point of combustion, and hence to qualify for 9-figure sums in annual "renewables" subsidies despite emitting more CO2 than in the days when it was burning coal (and vastly more than if the same electricity was generated instead by gas) with the distant prospect of maybe that CO2 being maybe absorbed by replacement trees (maybe) 50-100 years hence (maybe).

They've got me started now.

Anyhow, all four episodes are labelled "Series 1" so perhaps they'll follow up with more later.  There's no shortage of energy companies with big PR budgets, interesting stories to tell, and photogenic kit to display.  One thing we may predict: there will be a lot of Greens who are furious at the easy, glossy ride the Beeb has given the industry in these programmes, and will be pressing to get more *balance* into any subsequent series.



Anonymous said...

The submarine effect?

-- EC

rwendland said...

Something as modern as Windows 98 running an AGR nuc ND? Surely you meant to type "Ferranti Argus" ND - a 1960s to 80s industrial control computer which various sources say controlled Heysham 1.

There is a wonderful quite detailed IAEA paper about how Argus ran/runs Torness AGR, sister to Heysham 2. It used 46 Argus 700 computers with 76 processors: each reactor in the dual reactor station had 10 input multiplexing computers, 11 control dual-processor computers, and a supervisory triple-processor computer with a standby backup.

I'd hazard a guess they are still running them as I've never read about a major control system upgrade for any AGR reactor, and if they did modern standards would make the software upgrade of a Reactor Digital Direct Control system very expensive and time consuming. Shame the TV program didn't film some 40ish-year-old computers running Heysham, but I bet the PR folks told the manager not to let the cameras anywhere near a computer room!

Anonymous said...

If they are using kit that old, then the hackers must be even older. Or are the systems hack-proof?

Overall I think we have to recognise when it comes to power engineering, we ain't half bad.

Don Cox said...

I don't see how you could hack a computer that has no internet connection and, if it uses removable disks at all, will be using eight inch floppies.

There will be tape recorders for data storage.

The only way to hack a nuclear power station is to infiltrate a sleeper spy.

Don Cox

E-K said...

I've been in a helicopter on a nice day and even then I shit myself.

Too many moving parts and a feeling of hanging. Couldn't wait to get on the ground again and I LOVE flying - especially in a paraglider.

Bunjee jumping was good too. I also shat myself on a sheer rock climb at over 1000 feet in Morzine last year and froze but when the guide came back and said he could call a helicopter-and-winch out I soon got moving. (I also realised I'd booked the event without valid insurance.)

E-K said...
This comment has been removed by the author.
E-K said...


By now we must realise that it is no longer about giving us energy.

I have now succumbed to the Great Reset conspiracy.

Plebs like me were never meant to be in helicopters, on bunjee cords, rock faces in exclusive locations in France... unless on military service. They don't even want me to have a car or a boiler.

Does the insanity of these policies make more sense to you now ?

Anonymous said...

Tell us what you really think of Drax, ND.

dearieme said...

A non-woke series on the Beeb? Someone's taken a bung.

Anonymous said...

I just watched the nuke (Heysham) episode. Why do these people say nookiller?

rwendland said...

Historical addendum to Heysham's computers comment above: Decided to double check my history above, and it seems right despite not being able to track down a modern definitive statement.

To maintain these very old computers, you'd expect there to be a coven of old computer engineers supplementing their retirement with consultancy. Nosing into linkedin you find at least one: "Support Legacy Ferranti Argus Hardware used in PMS Systems in EDF Energy Nuclear Power Stations (Aug 1995 – Present)". And there is even a current Thales advert for Ferranti Argus software engineers with nuclear power experience for a 10 year support contract!

I did track down one Argus replacement project, but at the oldest AGR Dungeness B - the one that famously took 20 years to build and get working. The replacement cost was a whopping £75 million in 2015, and a regulator report says it replaced the "original equipment installed during the construction of the Station", so this was the first replacement. The first Argus was delivered to Dungeness in 1970, so this replaced computers that were up to 45 years old. I would guess EDF wants if possible to avoid this cost again for the a bit newer AGRs, so is keeping their Argus systems going.

Also the new hardware simply "replicates the functionality of the previous system", so the software saw minimal change - probably to minimise verification. If I'd been on the project I'd have looked hard at the option of simply writing an Argus 700 emulator to run the old code on modern hardware to tide over to the closure date.

Incidentally, these Argus systems did not do full computer control. Mostly for analysis of instrumentation data to present to operators and raise alarms. Also limited control akin to a 747-era autopilot, such as small control adjustments to keep pressure and temperature stable, and sequencing the complex/tedious startup, and possibly shutdown.

So it does look like most of our old AGRs still use ~40 year old computers! A difficulty EDF would have if it wanted to replace them is that the AGRs in the fleet are not all the same, and in fact some are quite different (it was much worse in the first generation Magnox fleet). No British company was big enough to build them all, so multiple companies were used, and they all tweeked the design, especially outside the nuclear core like control systems. So they cannot roll out one computer replacement design cheaply across the fleet, but would have had to do multiple ~£75 million replacement projects - ouch.

Nick Drew said...

V. interesting case study, Mr W!

20 years ago I was providing an energy software product (not nuclear control systems ...): our main product was big, 'enterprise s/w' with a very high exit barrier

and it worked! - and (more or less) still does, with upgrades along the way

as you say, "you'd expect there to be a coven ..." and indeed there is: the company has long since disappeared, but several of our guys still make a living by knowing how to (e.g.) update the static data / interrogate the deep archive / change the report-writers / figuringout the occasional upgrades - all the stuff for which the clients ought to train their own 'super-users' internally, but never do

rwendland said...

ND, yes its amazing how many good old systems carry on away from the flashy web interface. British nuclear was lucky the Ferranti Argus was a rather reliable machine, with a military Argus M700 variant (which I would not be surprised if still in use in some dark corners, like nuclear subs). The Ferranti Argus was in fact a quiet British computer success story, with many hundred (maybe thousand+) produced of the various models.

I have found another great source that explains the Argus AGR history much better, a 1996 Nuclear Electric short paper to an IAEA working group. The first 4 AGR stations, Hartlepool, Heysham 1, Hinkley Point B and Dungeness B, used the older Ferranti Argus 500 originally. These were getting long in the tooth by the 1990s, and all except Dungeness B were upgraded to use the more modern Argus 700s. But unfortunately the 700 has a different instruction set to the 500, so they got Ferranti to change the instruction set electronics so they could run the old 500 code unchanged - they call it an "emulator"! I think the I/O sub-systems were the same, so the bespoke external interfaces could be carried across, so this made sense. So these three AGR computers might be as new-ish as ~1985.

Dungeness B unfortunately had used the very fastest 500, and the 700 "emulator" was not fast enough to run its system. So in 1996 Nuclear Electric placed a contract to work out a long-term maintenance strategy. And quite possibly that strategy led to the £75 million project to replace the Dungeness B Instrumentation and Control system in 2015, described above. A 19 year project seems about right for the infamous Dungeness B! So at least Dungeness B computers were a special problem, and the Argus's elsewhere can trundle on to the end.

The other oddity was Hinkley Point B computers which were GEC M2140s with 64 kbytes of memory (ZX Spectrum size) rather than Ferranti Argus. These unfortunates were a less capable and unpopular computer, and Nuclear Electric contracted Ferranti and Ada Core Technologies to replace them ~1997 at a cost of about £25 million. Not sure if they used Argus 700 for the replacement.

NB The Science museum has a 1974 Argus 500 in storage, which ran the BP’s Field Control Centre in Aberdeen monitoring all flows in Cruden Bay and Grangemouth fields, which the museum grabbed in 1988 when BP upgraded. They have 700 parts and manuals, but not a complete system. Hopefully they will pick up an ex-AGR one over the next 10 years!

C & I REFITS ON AGR's - R Clarke, Nuclear Electric PL

Hinkley Point B GEC M2140 replacement