First puzzle is the price however - according to NextBigFuture these things are rated at 70 MW thermal (dropping to 25 MW electrical) which means that Hyperion seems to be pitching these things at ~$1/W of rated electrical output. Once you factor in the shorter lifecycle that seems somewhat more expensive than what a full scale nuke would cost, but not outrageously so. What am I missing here?
WRT capacity/scale issues - wikipedia tells me that oil sands post-processing uses ~1.25GJ in energy to produce a barrel of oil, which is equivalent to ~0.35MWh. One of these units will produce (70 * 24) 1680 MWh per day of thermal energy, which translates to 4800 barrel/day/unit. Call it 4000 barrel/day to account for capacity factors. The 2006 production average of 1.25 million barrel/day would therefore require a fleet of about 300 of these units (for a total price tag of $7.5 billion over a decade or so - which seems eminently doable by the oil majors I would say).
Future expansion would require more of course.
AFAICS the two principal advantages of the nuke battery over a centralised conventional nuclear power solution are (i) timeliness and (ii) co-location efficiencies.
Timeliness - the first of these nuke battery things could be on site in 3-4 years if Hyperion are to believed. Which means that they will be approaching the end of their life at about the time that Lac Cardinal is projected to come online in 2017. Thus, if central generation proved to be a better solution then a nuke battery fleet could still be used as a bridging technology until the central solution comes on line. In which case it would be displacing NatGas as the energy source for steam formation, which Hyperion claims a 70% efficiency advantage over (dunno if that is includes NatGas infrastructure costs however).
Co-location - oil sands require steam which can be produced directly by the thermal output of a colocated Hyperion battery farm rather than piping in electricity (or NatGas) produced offsite in order to manufacture steam. That's going to give you something in the order of a 50% efficiency headstart over an offsite energy source. This won't work for Shell's oil shale process of course.
Finally the small size and short lifecycle of these things, whilst a fairly significant handicap in terms of the scale/capacity issues that dominate the electrical generating biz, could paradoxically confer a valuable flexibility premium for the oil biz - which is famously gunshy about investing in long-term fixed infrastructure for unconventional resources. Being able to flex your battery fleet up and down on a year by year basis or redeploy your battery fleet as fields peak and decline (there's nothing that says these things have to be buried) is likely to be something that the oil biz will be very keen on.
Regards Luke -- #include witty_sig.h
Oil sands is a completely different beast than oil shale, and these things could be useful there. But I'd imagine that you'd need a lot more than a few 10's of MW's to power an oil sand operation. Probably 100's of MW's (I've recently visisted a paper mill which needs 350 MW power and 200 MW process heat), and then the pebble bed reactors are probably a better idea than fleets of nuclear batteries, if they ever get off the ground.
Possibly even full scale reactors might be the preferred choice here, they don't have to be 1000+ MW you know. And I don't think we should discount the fact that any plant built by AECL will have a certain attraction to local politicians. Peak oil is not an energy crisis. It is a liquid fuel crisis.
The question becomes whether Hyperion (and any other licensees) can meet the volume/timescale ramp up they are talking about in the press release and still hit their price point. Another advantage small unit size of each battery module confers is that this claim can be empirically tested with the sort budgets and time horizons that conventional corporate organisations are comfortable working - rather than requiring a govt get involved and take a multi-decade punt on the technology working out.
