Saturday, 23 March 2013

The bizarre economics of Hinkley C

This week, UK government approved the development of the twin 'Hinkley C' nuclear power stations by the French company EDF. It’s interesting to compare the economics of Hinkley C with an something radically different but equally effective.

Hinkley C will add 3.2 gigawatts (GW) of generating capacity to the UK energy budget. Because old coal-fired power stations are being shut down, there will soon be a shortfall in the budget, and so adding capacity seems like a good idea. Hinkley C will be costly to build: the cost is estimated at £4,375 per kilowatt of capacity. And though when we’ve got it, we won’t have to spend much on fuel, it’ll still have some annual operating costs; my estimate is that they’ll be £16 million or more.

An alternative project would be to get all the filament lamp bulbs in UK buildings replaced with LED lamps. Though they are small, they are many, and taken collectively the change would remove about 3.2 GW of demand from the UK energy budget: by a happy coincidence, that’s the same power as Hinkley C. But even replacing our estimated 400 million filament lamps would cost less than a fifth of Hinkley C. The cost would only be about £557 per kilowatt. And every year, that would give us annual operating savings of around £709 million.

LED lamps aren’t dangerous in an earthquake, or (so far as I know) in the hands of a terrorist, and unlike Hinkley C, they don’t give the UK a long term dependency on another country, and they don’t cause health scares.

We probably do need Hinkley C, but instead of the next nuclear power station after that, UK government really ought to have a go at the LED alternative.


Hinkley C
LED project
Quantifiable costs of acquisition

Unit cost of acquisition
£4.15 [6]
Number of units
0.4 bn [4]
Overall cost of acquisition
£14 bn [1]
£1.66 bn
Cost of publicity campaign
£146 m [11]
Total cost
£14 bn
£1.806 bn
Power per unit
1,600 MW [1]
54 [2]
Total power
3,200 MW
3,240 MW
Cost per kilowatt
Quantifiable costs of operation

Cost per kWh
-2.5p [9]
Cost per year
> 16m [10]
-£709 m
Effect on security (national and energy)

Dependency on offshore companies
Substantial long-term commitment
Short-term engagement only
Increased terrorist risk
Increased risk of serious accident

Thanks to George May of Métamorphosen, 'Nic Newey of IPL and Anthony Price of Swanbarton.


[1] Wikipedia, Hinkley Point C nuclear power station

[2] Replacing a 60W filament bulb with a similarly bright 6W LED bulb [3] would save 54W

[3] Wikipedia, LED lamps

[4] 0.6 bn lamps in UK [5], of which ‘most’ are filament type (my conservative estimate 0.4 bn)

[5] BBC, The 60W bulb: a luminary love affair, 2011

[6] BIC E14 6W bulb, 2013

[7] 21,600 MW if all the bulbs were on all the time. Assuming just 15% duty cycle give (15% * 21,600) = 3,240 MW

[8] Royal Academy of Engineering, The cost of generating electricity

[9] Typical cost from [8]

[10] Operating costs for Hinkley C not published , but it’s widely reported that it will ‘provide’ 800-900 jobs in the long term. Assuming a cost per employee of at least £20,000 p.a., that’s at least £16m.

[11] Cost of publicity campaign would be comparable with the cost of the recent UK Digital Switch-Over publicity, which was estimated at £146m.


jmdesp said...

I'm not sure the numbers are precisely correct, as the wikipedia page says household LED currently generate around 64lm/W which means they are not really doing better than CFL. So according to the Cost Comparison table, you need a 13W bulb to substitute for a 60W one, with only 47 watt of savings for 10$. This doesn't increase the cost that much, but means that the number of bulb to replace grows to pretty much every single bulb instead of only 400 millions.

And this gets us to the crux of the problem, the load factor. Hinkley will be around 90%, which represents 21.5 hours a day.
Reference are an average between 2 and 3 hours of use of a lamp per day, so between 7 and 10 times less. If you replace every single lamp instead of focusing on the most used ones in your living room, it will go in the lower range.

This ends up making your LED alternative actually more expensive than Hinkley, except maybe if focusing on the lamp that are the most used, but in that case the scale of the power saved is much lower. There's also the rebound effect, as the consumption from the LED bulbs will be very low, people won't feel motivated to switch them off when leaving a room, which leads to more consumption.

Nonetheless I'm all for more LED use, especially since the highest use is at peak load, in cold winter evening. But it should not pass for the game changer that it's not. CFL are already sold since quite a while, and are a significant part of the bulb used (especially the ones frequently used), and not much of a change was seen.

Clive Tomlinson said...

Many thanks to jmdesp for a thoughtful and informed comment.
I'm not sure that I agree about the load factor. My blog stated a guessed duty cycle (load factor) of 15% for the LEDs, which is very close to the 2-3 hours that jmdesp's unspecified 'references' give. And I'd assumed 100% load factor for Hinkley C, which is crude, but downrating that to 90% only makes the LED argument stronger.
Anyway, I don't want to quibble. The point I was making was about shifting attention from supply-side expansion to demand-side shrinkage.
I'm typing this by natural light, in one of the UK's rare hours of sunshine this year.

Unknown said...

Wikipedia now gives the cost of Hinckley Christmas as £18 billion.

I recently bought 9W LED lamps to replace 60W filament lamps for only £4 each.

Costs and technology trends continue to move against Hinckley Christmas.

Clive Tomlinson said...

Six years later, the BBC has caught up with us:

Post a Comment