Re: @Steve Todd & Paul Crawford
"Now then, where are the GW sized generators, and is anyone prepared to dig up towns and cities for hot water distribution to make said efficiency worth it?"
You've hit the nail on the head, that the cost of district heating is too great in most cases in the UK, with the exception of high density housing. The fundamental cause of this is that the cost of heat distribution networks is too great. Heat distribution pipes cost around £1,200 per metre of length in real life conditions*, and that soon mounts up (even for terraced housing this typically means £4k of property specific heat network, before you've built your energy centre (say £2k per property served), or installed a heat interchange unit (HIU) and heat meter in the property (£2k). This compares to less than £2k for a gas combi boiler fitted and £600 of property specific gas network and perhaps £0.5k/property of upstream gas production gear.
If you've got a number of close together apartment buildings then a small area district heating system makes sense serving a few thousand apartments, but nothing on the scale seen in (say) Sweden or Poland, where winters are colder, average space heating demand much greater, and there's no national gas network like in the UK.
Then we come to the alleged efficiency benefits. In principle they exist - you run a CHP engine (typically gas fired spark ignition reciprocating engine) for heat and capture electricity as a by product. Problem is that the heating demand varies year round, and during the day and it's simply not economic to build a CHP that is mothballed for the six months of the year outside the heating season. Then, within day the heat load varies dramatically. Overnight a well designed system can coast on the thermal inertia of the system, but during the day you have to serve morning and evening peaks. If you run a CHP engine to serve that load profile then (a) you've again got an asset sitting around for three quarters of the day twiddling its thumbs, and (b) you have a problem with the reliability - regardless of what the makers claim, power generating sets respond badly to excessive cycling, and we've found that running the CHP on simple double cycling per day meant very noticeable decreases in reliability (which means higher maintenance costs and loss of use). You can build additional heat storage into the system to allow the CHP engine to run for longer, but the problem is that "storage" in this context is a bloody great insulated hot water tank, and when you build these at any scale they become very expensive. I work for one the largest operators of district heat in the UK and another northern European country, and we do know what we're talking about.
So, what this means is that for a UK heat system you size your CHP for your year round baseload, and all of your seasonal and daily peaking is delivered by gas boilers (sometimes biomass assisted, but there's lots of reasons to not want that). You could use all CHP, but the costs would be astronomical, and you have to remember that the system needs to be wildly over-provisioned against average demand, because you have to allow for a plant breakdown and maximum demand in the coldest winter conditions (say -15C). So a small scale district heat system serving a few hundred properties would have a 185 kW electricity/230KW heat spark ignition CHP that runs between 05:00 and 23:00 every day of the year other than when down for service. You then have say three 3MW boilers which would never be used all together - two of them can serve maximum heat demand, giving some leeway against severe winter demand, reduced output for lower mains gas pressure and loss of both the CHP and one of the boilers. This will be why the university is extending using boilers - unless you want to generate electricity at well above the cost of grid power it doesn't make sense. Over the year the electrical power output is around 5-10% of the total energy output, so even as a by product it isn't making much difference. This is in strict contrast to industrial CHP where the plant is typically a 50 MW gas turbine that runs at constant load, and produces 30% of power output as electricity, and the balance as process heat (with often incidental delivery of space and water heating).
So in aggregate terms, the overall efficiency benefit of district heating CHP units is much lower than proponents claim, unless you want to chase operational efficiency at a very high cost indeed. Despite this the clowns at DECC are earnest believers in the benefits of district heating, and are pushing developers and power companies to roll out heat networks - like wind and solar, very expensive non-answers to the question of climate change. Over the past couple of years DECC's heat network delivery unit has been handing out a few million quid to the beards & sandals of local government to encourage the take of heat networks, and they in turn believe this simplistic and inaccurate mantra that district heating is more efficient than centralised electricity and local heating systems.
Having said that the efficiency is a dream and the costs are high, I would point out that in high density housing it can be slightly cheaper than heat from an individual gas boiler (not permitted in high rises anyway), although the costs are very different - the operating costs are lower, but the capital costs are a lot higher. In functional terms, district heat works as well as gas: It is as convenient as a gas system, safer, requires minimal in home servicing (typically a fifteen minute check over of the HIU once every two years). Many middle class people in the UK have a problem with district heating that you can't change your supplier - the energy service company that owns and operates the heat network has you as a captive customer essentially for the life of the property. Residents in social housing tend to me more accepting because they're often used to communal heat delivery at a building level, or their alternative is crap and expensive dry electric heating systems.
So there you have it. Ledswinger's District Heating Primer. Conclusion: Works well for high density housing where you'd normally use a communal boiler or dry electric systems, but otherwise a very expensive idea that has few real world efficiency benefits.
* Heat pipes are expensive because circulation speeds are low to achieve efficient heat transfer and avoid high pressures that increase pipe bursts, so you have larger pipes than for gas or cold water distribution, usually steel for durability, and you then have a good thick layer of insulation. This means you're laying a more expensive pipe than MDPE gas or water pipe, and a much larger pipe, leading to greater civils costs. There have been experiments with "4th gen" heat networks using low temperature distributuion, meaning less insulation is needed and heat losses are lower, but this requires the house to have a heat pump on HIU, which increases capex, makes the device more complex, and increases the electricity bills, so its a bit of a zero sum improvement.