Before I get to the data centre reasons I might as well fess-up as to all of the other reasons that I voted to ‘leave’.
First of all, it was the first time I had voted in any election for over 40 years, something, given the huge turnout for the referendum, that I clearly shared with a lot of people. In fact, there is a link connecting my usual apathy with my keenness to vote this time – I have a very low regard of politicians of every persuasion, gender, creed, religion or nationality.
For me, the professional Civil Service runs the country well enough and a marginal majority in the House of Commons prevents the government of the day from doing too much harm. I will refrain from telling you what I think about the House of Lords. Let me point out, before you think that I am wasting the franchise that so many people died to give us, that no matter how I had voted since the early 70s there would have been no change in my MPs affiliation.
Our lack of proportional representation means that 25% of the electorate can and does elect a government with a healthy majority. Hardly democracy. The 39% of the electorate that voted to leave (more people voted to leave than have ever voted for anything else before) would produce a landslide the likes of which we will never see.
I certainly didn’t vote ‘leave’ because of immigration (my father was one) or any anti-European feelings, since I revel in our social and historical diversity and most of my friends are not British. No, I voted so that I could show what I thought about UK and EU politicians warning us that we would be stupid to vote leave. I voted to protect what democracy we have, to avoid an EU Federal State and Armed Force and to object to the excesses, waste and gravy-train in the EU institutions. And when the CEOs of two German conglomerates tried to frighten us off I think that it contributed to tipping the balance at the last minute. As for the single market – we are the fifth largest economy in the world and we buy far more goods ‘in’ than we manufacture and sell ‘out’ – they will be queuing up to sign trade deals with us, not the other way around. We had a single market before the EU, called EFTA, and that worked just fine.
If you want to be cynical you could explain my vote as simply a combination of my age, gender, ethnicity and that fact that I don’t live in London.
But what about data-centres and the EU? Well that’s a sorry tale that would have ended worse with us ‘in’ as ‘out’.
Having spent a lot of time in and around the offices of the Commission I have always been frustrated by the same background noise; ignorance. When the EU wants to do something it appoints a consulting firm to write them a paper and then, more often than not, gathers industry in and shows them the report. It is, as you would expect not an expert piece as the consultants rarely know the industry they are asked to report on. They then ask for industry input for one/two days while the chosen consultants (often staffed with bright young things straight from a Masters programme having bypassed a real-world job) scribble away. A few months later the ‘final draft’ is published and the industrial invitees are called together again to witness the birth. Hey presto! A poor job done at great tax-payers expense. The latest was an energy efficiency improvement document for data centres which started off with the invited audience asking why the consultants hadn’t even referred to the EU’s own Code-of-Conduct. They hadn’t even heard of it.
Then came along the Emissions Trading Scheme (ETS) designed to record and tax carbon emissions from small power stations or businesses self-generating. Noble, yes. Well applied, absolutely (stupidly) not. The EU put all generators (over about 7MW capacity per business entity) into the scheme even if they were standby emergency sets. This means that a data centre business may have to spend £30,000 per year in administering the reporting via a consultant and pay £1,000 in tax as a result. Same applies to a large hospital. No amount of objection worked, in fact they virtually refused to discuss changing it.
But the waste of tax-payer’s money on data centre related ‘research’ has a much longer history. It goes way back and includes the grants emanating from the EU FP series of funding, i.e. FP6, and the last round FP7. FP7 is spending €50 billion on research projects and the EU has spent multi-millions of those euros on such projects as GAMES, FIT4Green, All4Green, CoolEmAll, GEYSER – even PEDCA and Eureka which have been centred on UK co-funded groups.
Why do I call it a ‘waste’? Well, with the jury still out on PEDCA (a study of training and skills required run by our own DCA) and Eureka (smart city integration of data centres focussed at UoEL) the rest have been paid for, finished, printed and put on a shelf to collect dust. They will never see the light of day and only exist in the memories of their authors in terms of a meal ticket for a couple of years. Much of the research funds are spent on travel and accommodation when attending conferences.
No doubt that research should not only be regarded as a success if it finds a solution to something but many of these projects have been ill-conceived with obvious and predictable results. One, for example, explored the possibility of matching data centre load with the availability of intermittent renewable energy, like wind. It seriously brought into the arena of social networking like Facebook what amounted to batch-processing. So, when the youth of tomorrow opens Facebook a message may pop-up saying ‘come back later when the wind is blowing’. Now, don’t get me wrong, that is not a bad idea but it takes two minutes to think about (and probably reject as political suicide) and should not involve a couple of million euros spent over two years by several people.
The same with the re-use of waste heat. Clearly a great idea in small embedded data centres of the future and technically no problem once you decide to go fluid-cooled. Most large city-centre residential and hospital/elderly-care buildings can absorb 100-200kW of hot water at 65C on almost continuous basis. Again – quick to propose, easy to implement using technology already available and few hours of work to write a decent white paper on the subject, but a million or so euros and two-three years for a consortium of several companies/universities? No, it is akin to robbery (without menaces) from an ignorant bureaucratic organisation that should either employ its own staff or stick to politicking and filling in expenses forms.
Then the whole issue of renewable energy for ICT – the focus of so much research money under many guises. The EU is obsessed with the idea that ICT should be renewably powered. Even when faced with the question in public ‘why should tap-dancing dogs (YouTube) be renewably powered over a hospital?’ they continue to voice that ICT should be a modern exemplar for all industries. I even tried to compare ‘pornography over income-tax collection’ but to no avail. Yes, of course, all energy should be renewable, but it is never going to happen without a nuclear base capacity or the collapse of western society.
There is even one co-funded project that was set up to coordinate the others since most of the projects actually have the same core topic. That report should be worth reading.
The only project not funded (yet) is the one that would propose that phones can’t have integral cameras and that cameras could not be internet enabled. The drop in internet traffic (and power) would be near to 40%. There, one minute to type – where can I send the invoice for €100,000 of the tax-payers cash?
The last frustrating data centre related point is that of energy reduction regardless of the criticality of the load. I have been in open discussion in Brussels trying to make the case that PUE should be related to the data centre business model. In other words, Google might be 1.12 but a bank might not be prepared to go as far down as 1.3 as it involves ‘risk’, either real or perceived, to the availability of digital services. In addition, let’s not forget that a bank’s energy cost in its data centre may be as much as 0.01% of its turnover – turnover that is driven entirely by the digital services provided by the data centre. Those numbers drive behaviour and I have found much resistance in Brussels to even acknowledge that a PUE target is not practical unless business is put first.
During that discussion (in 2014), I was made aware of the sort of agenda that would have arisen if we were to have remained. There is no point pretending that we are liked. We didn’t join the Euro, we are outside Schengen and we like the Americans – what is there to like if you are a European Federalist? They don’t like the fact that 40-50% of Europe’s raised floor is in the UK since they feel it should be in the heart of Europe (although I have never worked out if Paris knows that that is Frankfurt). They are irritated by us when we tell them that no European equivalent to ASHRAE exists and Telecom Italia can’t re-write the rules without the ICT industry involvement – and that would be TC9.9 of ASHRAE. They wouldn’t accept that it is not the data centres that are driving power consumption but the people – and not (back then at least) business or medicine or education. Telling them that their 20MBs digital access agenda is driving the power consumption they are trying to reduce was not welcome – and trying to tell them who William Stanley Jevons was fell on stony ground.
But the killer comment I heard involved ‘the London market’. Thanks to my immigrant German speaking father, I can listen well enough in German. Thanks to my English mother, and growing up in the very early fifties when German was not a popular language in London, I speak only English. So, it has always been useful not to make it obvious that I am anything else than a typical Brit – when speaking to foreigners we just raise the volume. That’s when, on this occasion in 2014, I overheard a secretive discussion in German during a coffee break – that the data centre power problem would be easier to solve once the main financial market was moved out of London and into Frankfurt.
So, it is not enough to allow us to be a buyer of goods rather than a manufacturer (and become an 85% service based economy) but we also have to relinquish our finance, banking and trading business. Now we have to fight for it.
A last question: If all the other EU members had a leave/remain referendum how do you think the electorates would vote? I think that our result would be repeated in well over half of the member states.
Ian Bitterlin is a consulting engineer, visiting professor at Leeds University and regular contributor to Mission Critical Power.
Ian Bitterlin: Data centres being singled out by EU over energy efficiency
Let them eat bandwidth: Ian Bitterlin on optimisation or bust
Click here to see if you qualify for a free subscription to the print magazine, or to renew.
Follow us at @mcriticalpower. For regular bulletins, sign up for the free newsletter.