You may have heard rumours that the Clover experiment has been cancelled, and indeed Peter C has just written about it. My colleague Alan Heavens had some observations on this turn of events, so I invited him to write about it. Here is his piece :
A few days ago STFC cancelled the Clover project – a casualty of the financial crisis in the council, and perhaps not the last. What should we make of this?
First, some background. Clover was a project to detect B-mode polarisation in the Cosmic Microwave Background, would have detected the gravitational lensing effect of intervening structure on the CMB E-mode polarisation, but much more excitingly it was designed to seek evidence for inflation, which in the standard cosmological model is assumed to have driven the expansion soon after the Big Bang. This is noble and potentially Nobel science.
Clover’s cancellation raises questions about which projects the UK should be funding, especially in lean times. STFC’s stance on this at the beginning of the crisis was set out very persuasively by John Womersley in a talk in Edinburgh and elsewhere – the UK should be pursuing projects of high scientific importance, with big international impact, and with UK leadership. The message was clear – we should do fewer, more important things, and lead them. Absolutely right, I thought – this is what our research councils should be saying.
Clover ticks all these boxes – it was purely a UK project; a positive finding would have been of colossal scientific importance, and the impact would have been enormous. So what went wrong? The bare answer is that the costs went up, and STFC Council reconsidered it, and cut it. The project certainly faced challenges, in detector technology, and in the team moving from two institutes to four, but the project was well advanced. Unlike the VSA, where an opportunity to fund a seminal experiment well in advance of competition was missed, Clover had been funded early and boldly. The delays meant that it might not have been the first to an accurate detection, as other experiments were catching up, but it still might have won the race. The fear of competition can be damaging; other projects are not immune to delays and difficulties and one might miss an opportunity. A salutary lesson comes from 2dF and SDSS: had PPARC taken the same view of 2dF in the face of the published schedule of SDSS, the important scientific results made by 2dF in large-scale structure would have been lost to the USA.
As always the case when finance is the issue, the decision is a matter of scientific judgement and strategic priority. On finance, it would be interesting to know if a minority partner was sought. On scientific priority, then the advisory panels must be allowed to advise and the Science Council allowed to decide, but it would be interesting to know what advice informed the decision.
The Clover shortfall is, incidentally, the same figure as is rumoured on this blog to have been allocated to MoonLITE.
Just a quick note for those interested in this matter: Physics World have run a news item on their website and there will be a story in next week’s Nature too.
The Physics World item reports that CLOVER was almost 60% over budget. Without wishing sound sceptical at a time when the harsh realities for many people’s jobs looks poor, if true, doesn’t this sound like rather poor Project Management by both the Project Team, and presumably by an Oversight Committee (which I would assume is the responsibility of STFC)? Most well run projects have a contingency allowance that should take account of the normal inevitable variance on costs. But with such an over run, could CLOVER really have been expected to have retained credibility in a harsh funding environment when it over runs its budget to such a large extent – how did the Project Team possibly get their finacial planning so wrong – can they advise others on how to avoid this in the future? For the meantime I’m sure that as a community our deep sympathies all lie with people whose jobs are threatened. It’s a sad time, not helped by the actions of STFC Executive over the last two years.
From the figures Peter quoted, it sounds like more like 30% over budget, but the point remains. With the best will in the world, sometimes projects simply do cost than you thought. The contingency thing is just guesswork. For something run-of-the-mill, when this happens, the right thing to do is to cut your losses; but if you really believe something is ground-breaking, you swallow hard and keep going. Its a tough judgement call. The original SCUBA project went considerably over budget, but keeping going turned out to be right, as it broke completely new ground for the UK – others are still catching up. (I will resist listing specific counter-examples of expensive white elephants…)
I won’t comment on the reasons for the cost overrun (it’s complicated and I don’t know the details because I’m not privy to the details having joined the project in a minor role relatively late in the day). The saddest thing, though, is that the equipment is pretty much ready and the extra funds were needed to ship it out, set it up and run it. The decision to scrap it therefore wastes all the work done so far.
Such decisions have always been difficult but when the funding pot is shrinking so drastically they are now becoming impossible. It really is a no-win situation.
ps. The extra funds needed are £2.5M on an existing spend of £4.5M. That is about 60% extra.
Andy — I guess STFC would argue that all their science is groundbreaking, and that when the budget is so squeezed that there is no system-wide contingency, then something painful has to happen in cases such as this. However, if the competing “science” really is the likes of MoonLITE, then we should continue to ask serious questions about STFC’s priorities.
The story of Clover is a tragedy for all those who have worked so hard on the project and for all of us who had such great hopes for it. It is important however understand some of the background to the cancellation of the project by the STFC Council.
Clover was originally approved by PPARC Council in 2004 at a cost of £4.78M. In 2007 an independent review put the expected cost to completion of a descoped project at £5.6M. Following two critical design reviews that year, STFC put in additional project monitoring and project management support, but given the high scientific priority of the project decided to continue with it. In 2008 a further review concluded that the cost to completion including contingency had risen to £7.54M. At that point it was decided to take the project back to STFC Council for a decision on continuation or cancellation. By that time the project had spent something like £4.3M. Council reluctantly reached the conclusion that despite the continued high scientific priority of Clover, it could not risk the considerable additional resources needed to complete the project.
We now need to work with and encourage the project team to see what we can rescue from what was always going to be a high risk endeavour.
A significant part of the extra money required for Clover is FEC overheads on people’s salaries. Clover was originally funded pre-FEC and so the original budget and the overrun budget are not directly comparable in terms of man-months effort or money to be spent by the project itself, as much of the overrun budget goes to the host institutions in overheads. Certainly the direct ratio of the overrun budget to the original budget is not a reflection of the amount of work to be done to complete the project.
Another reason for part of the increase in costs is that in the orginal Clover proposal, the instrument was to be sited at the new French/Italian Antarctic base at Dome-C, with the French and Italian Antarctic agencies funding the site infrastructure. However it became clear in 2005/06 that this base would not have the infrastructure required to support Clover on the project’s timescale, and so the site was shifted to the Chajnantor site in the Atacama, with the infrastructure costs to be funded by the project.
Clover always had some significant risks associated with the technology development – indeed the ground breaking nature of the technology was counted in its favour when the original funding decision was made. These risks are now mostly close to zero – almost all of the technology required to complete the project is now working, and the project is mostly in the early assembly and integration phase, prior to initial deployment in early 2010.
It is also very important to point out two facts about the cost increases in Clover.
First, the original cost did not include the site, which was hoped would be provided by collaborators. This fell through, and the site costs have had to be added to the uplift requested. STFC were aware from the start that this was a risk and that an uplift beyond the original contingency would be required in this case.
Secondly, while the cash numbers Richard quotes are correct, the original cost (4.78M) was pre-FEC, whereas the increases over that are calculated under FEC. The cash costs of staff are roughly doubled under FEC, so the increases look disproportionally large compared to the original cost. What has happened is that the project has taken longer than expected – not surprising give the large amount of new technology that has had to be developed.
The project is now actively looking for partners to allow it to continue in some form, and we certainly hope that STFC will support this.
F and Dave
I very much agree with you points and Council were made aware of these factors. In the case of fEC the unfortunate fact is that once the project had to go back for re-approval and any grants had to be reissued, the cost had to be calculated on the basis of fEC and regardless of where this money ends up, we need to find it from somewhere.
Somewhat late into this posting, I guess one of the advantages of Clover referred to by Alan – that it was a UK project – may have been its achilles heel. Had it been a multi-partner project, with legal agreements over penalties for withdrawal, STFC might have been unable to give it the chop so suddenly, and so late in the day. I suspect other UK-only facilities will be getting anxious..The latest Council minutes from Feb 09 refer to the potential need for additional £15m savings moving into the next SR period.
Paul raises an interesting general point. It is a slightly sad one, that UK projects might be advised to find partners (thus diluting UK leadership) in order to give some protection from their own funding council. It rather undermines the relationship of research councils and researchers working together towards a common goal.
Indeed, I was at an Oxford PhD interview a week or two ago, and Michael Jones took me round the basement to show me all the kit. Both scopes are completed, and one mount is fully welded complete with hugely expensive ‘squashy gear’ motor. The mirrors are in place, and covered, and a space had even been cleared for it to be assembled and tested in the basement. On my frequent trips to Cavendish, I think the TES detectors are mostly completed too. It’s annoying to be able to handle the kit yet know that it will never make it the Chile!
Fingers crossed the Planck rocket doesn’t blow up 🙂
It does help to be in international projects with commitments. WFAU and CASU (data centres in Edinburgh and Cambridge) were ranked low by STFC but we were saved by commitments to ESO (VISTA) and ESA (GAIA). I think STFC were wrong in their ranking anyway, but I think good advice would be to be involved in international projects if you want to avoid being cut later. When I worked in the US, having overseas collaborators could sometimes be a disadvantage as NASA and other funding agencies saw overseas collaborators as a risk that they couldn’t manage – they could not exert any pressure if delays occured.
The WFAU/CASU judgement was more explicit. It wasn’t that we were ranked low but “rescued” by political entanglement. The feedback actually said we were low ranked excluding the ESO and ESA commitments. But as those commitments are nearly all of what we do, I was always a tad puzzled…. Hey Ho.
For big projects at least, international partnership is often sought out by funding agencies; as well as saving money, its a proof of world standing – other people want that thing.
Not very healthy that “Made In Britain” actually puts you at a disadvantage though, is it? Jet engine anyone?
The ideal thing is to be led by the UK, but with international partners. I don’t really see a political issue here, or an issue of principle. Politicians and reviewers love the “made in Britain” badge, but also like saving money, and seeing our efforts validated by partners. This is all perfectly sensible. The problem here is only a practical one, which occurs as budgets tighten, as others have said. I am pretty sure that STFC would prefer not to cut UK-only projects, as they in fact score well politically; but in practice they may have no choice, as this is the only way they can save money. This obviously applies to Merlin, and to UKIRT, which is scary. In principle it also applies to Diamond. STFC won’t find takers for international buy-in to Diamond, precisely because every other nation also wants its own synchrotron as a badge of pride. But in practice the “made in Britain” thing for Diamond is so huge, it is impossible to back off.
There are two rather obvious reasons why funding agencies favour international collaborations. The first is that it allows us to take on projects that we just couldn’t do on our own (LHC, ALMA, JWST). The second related factor is that sharing costs enables us to fund (and you to do) a far wider range of science. That said UK leadership is a key factor in assessing the impact and therefore priority of projects. Clover scored very highly on this of course. The problem with UK only projects is not that they are easier to stop (though clearly this must be true), it is that there is no easy way to look to partners for help in tough times. If you haven’t got partners, it’s too late to look when the project needs extra cash.
We do have a partner in Diamond of course in the form of The Welcome Trust with their 14% share. Not international but with many of the same features. And John Denham in a speech just a couple of weeks ago signalled that that any new light source would need to secure international collaboration.
‘UK leadership with international partners’ sounds a good description of the SAFARI instrument for SPICA. And yet STFC have said that UK leadership has to be given up for this instrument despite highly rated science, a well established UK community in this field and clearly productive and pro-active leadership of the project so far. Something just doesn’t add up here.
Peer review eh. Can’t live without it, can’t control the outcome.
Sure – when it’s as opaque as the system we’re stuck with.
You mean peer review like the process that Science Committee went through with MoonLITE, firmly recommending that it be placed at low priority. Remind me how much STFC has decided to invest on the basis of that advice.
Hi Andy — wasn’t really intended as a political point, nor even a dig at STFC who are simply playing by the rules with which they are presented. Rather more a philosophical one that there is surely something awry with a system that is skewed toward supporting a lower-ranked international collaboration over a higher-ranked domestic programme.
£1.5M for a Phase A study as I presume you already know.
Watcher – whose peer review is better – ESA’s or STFC’s, since the results wrt. SPICA/SAFARI have been vastly different?
I’m unsure of the figures, but 1.5M for Phase A for MoonLITE sounds larger than the Phase A funding for any of the cosmic visions programmes in the UK. That really isn’t sensible, especially for an area of science where the UK has no standing community or long term involvement.
[…] may also like to read the article by Alan Heavens over on the […]
I have been working for Clover since 2005. My role included the Site & Logistics organization.
First fact: I negotiated with the French and Italian Polar Agencies years ago to obtain logistic support at Concordia, the French-Italian station on the Antarctic Plateau. This was done through an agreement with the Rome University group (the same that run the Boomerang experiment in Antarctica a few years ago) and other French groups. I have the proposal accepted by PPARC (in 2003) in front of me. The last statement of the proposal Executive Summary says:
“The project … will cost PPARC £4.6M including VAT, salary uplift and working allowance up to and including commissioning, but excluding operation”. So, it was clear from the beginning that operation costs were excluded for the original budget.
Second fact: the negotiation with the French/Italian failed because of problems with the Station management that were out of our control. Essentially, the French and Italians, despite the collaboration with one of the leading Italian Cosmology group , and the evident outcome from the project at a very low cost, decided not to support a “British” project.
What I would have expected from STFC after years of extenuating reviews, was a proposal like “you asked for 2.5M£. Unfortunately, we don’t have them. What could you do with, we say, 1M£? Could you negotiate the support with another group to cover the remaining part of the required operation budget?”.
STFC abruptly announced the cancellation of the project, without any further possibility to save all the hard work and hardware done during the last 4 years. This is like to pay for a car and, after discovering that petrol cost raised, tell “Sorry, we just don’t have any money to pay for the increased costs of the fuel. Throw away the car”.
I think that the peer reveiw results were somewhat different for BepiColumbo too. What do you make of that?
Watcher: I actually can’t recall what the BepiColombo results were. I know it’s a field the UK isn’t that involved with so there are valid arguments for ESA to be involved but not the UK. This is definitely not the case for far-IR/submm astronomy where the UK has historically been a world leader.
[…] leave that discussion to those who know more about it – good coverage of the situation over on Andy and Peter’s blogs and on Paul Crowther’s […]
Morgan Freeman, to do the wrong thing. ,
All the while, they make the rich white men that own them richer by the day. ,
Two odd uncaught spam comments on the same old post on successive days ? Somebody somewhere must be writing a PhD on how all this shit works.
I think Miss95 might be my friend Flora who’s trying to butter you up because she gets jealous whenever anyone mentions Clover.
Dear Mrs Trellis,
buttering up sounds rather exciting. Do young people these days still have Mazola parties, or is that all rather old fashioned now ?
Dear Mr East Ronomer,
I think people generally use olive oil nowadays, at least in the suburbs.
That’s what my friend Marge says anyway.
curses. wrong login. gave the game away.
Dear Mrs Trellis
too late. I have already leaked your identity to the Crawley Gazette and Clarion, who have paid me handsomely.
At least Mrs Trellis never claimed to have a PhD.