Emissions Nuclear

Carbon footprint of the Olympic Dam uranium mine expansion

My home state of South Australia is host to the single largest known deposit of uranium in the world (by some estimates, up to 40% of the verified global reserves, although uranium is still poorly explored worldwide). The mine that was  first established over 30 years ago to exploit this resource (as well as copper [majority of production], gold and silver), Olympic Dam, is run by BHP Billiton, following their acquisition of Western Mining Corporation Resources in 2005. In that year, production was about 220,000 tonnes of copper and 4,600 tonnes of uranium oxide.

Late in 2008, despite some criticism, Premier Mike Rann gave the go-ahead for a $7 billon mine expansion. The eventual production of the enlarged mine — perhaps the largest ever man-made hole in the ground — is anticipated to be 730,000 tonnes copper, 19,000 tonnes of uranium oxide and 25 tonnes of gold per year. Critics have pointed out that the carbon footprint [e.g., diesel from vehicles and mining equipment] and electricity needs [possibly via gas-fired power plants] of this expanded enterprise would be massive — by some estimates almost 700 MW of extra electrical power demand. Greens SA MLC (State Senate), Mark Parnell, an active sustainability crusader here in this state whom I respect greatly, was quoted as saying: “Our state risks being left with a huge carbon black hole as we become the greenhouse dump for one of the world’s richest companies“.

Here I do a rough, back-of-the-envelope calculation, to test Mark’s ‘carbon black hole’ assertion with respect to the uranium extraction. My carbon footprint calculations are based on the careful figures derived for the highly detailed report by Bilek, Hardy, Lenzen and Day: Life-Cycle Energy Balance and Greenhouse Gas Emissions of Nuclear Energy in Australia (2006) [henceforth BHLD]. If in doubt, check the figures in that document yourself — its authors include three of Australia’s top life cycle analysts from the University of Sydney, who were commissioned by the Department of Prime Minister and Cabinet. I should note that a formal environmental impact statement for the Olympic Dam expansion, which will take into account many factors including water use and localised impacts, is due to be delivered in May 2009.

The current production of uranium oxide is about 4,600 tonnes, composed of a mixture of fertile 238U (99.29%) and fissile 235U (0.71%). Light water reactors (LWR) require enrichment of uranium oxide to 3 to 5% 235U, and once operational for a few years and running at a 90% capacity factor, need about 29 tonnes of enriched U per year per gigawatt of electricity (1 GWe) generated by nuclear power plants (BHLD, pg 79). It works out that in 2005, Olympic Dam produced enough fuel for roughly 22 GWe of generation capacity using today’s reactor designs (PWR/BWR), or 192,200 GW hours per year (GWh/yr). By comparison, the expanded Olympic Dam, with its anticipated output of 19,000 tonnes of U, would yield about 94 GWe of average power supply, or 794,000 GWh/yr.

Once the full nuclear life-cycle emissions for LWR are accounted for (includes: mining, milling, enrichment, fuel fabrication, reactor construction and operation, decommissioning and storage of spent fuel), the greenhouse gas intensity of the power generated is 60 kg CO2-e per MWhe (BHLD, pg 172), which is 60 tonnes per GWh [for fast spectrum reactors like the IFR, it would be substantially lower, since we skip the mining, milling and storage steps]. Thus in 2005 the emissions equivalent for Olympic Dam uranium mining was ~11.5 million tonnes (Mt) of  CO2-e. The expanded mine will be 48 Mt — that is, an additional 36.5 Mt of CO2-e will be released into the atmosphere each year as a result of the Olympic Dam expansion.

Okay, let’s put that figure in two different perspectives.

In 2005, South Australia’s total emissions were 28 Mt CO2-e. Let’s say that, despite our best efforts, our emissions continue to grow, such that in 12 years time they are 30% higher than in 2005, at 36 Mt CO2-e. The lifetime emissions that result from the mine expansion will be about the same as the total emissions of South Australia. Sounds bad, eh? Well, not to the atmosphere, which is a global commons.

Let’s say that uranium wasn’t available, and those French and Chinese (etc.) nuclear power plants were shut down and replaced with fossil fuel plants. Here is the comparison for the 2005 and 2020 (expanded mine) output based on three alternative power generation methods that yield the same total GWh/yr:

Black coal (supercritical): 2005 = 181 Mt CO2-e and 2020 = 747 Mt CO2-e

Brown coal (new subcritical): 2005 = 226 Mt CO2-e and 2020 = 933 Mt CO2-e

Natural gas (combined cycle): 2005 = 111 Mt CO2-e and 2020 = 458 Mt CO2-e

So, under the best-case gas alternative, we get an additional 410 Mt CO2-e dumped into the atmosphere each year if the Olympic Dam output was cancelled. With brown coal, the stuff we find powering Victoria’s Latrobe Valley (and feeding SA via the interconnector), we get an extra 885 Mt CO2-e. That’s a whopping 37% more than the estimated 530 Mt CO2-e we might expect Australia to be emitting from ALL emissions sources in 2020, assuming we manage to meet the 5% reduction target of the CPRS. Or looking at it another way, the Olympic Dam expansion will ‘offset’ South Australia’s total carbon emissions by around 13 to 26 times.

For a counter analysis on expanding our coal production see here: Save a bit here, ship a whole lot there.

So, to conclude, I agree with Mark Parnell that “Our state risks being left with a huge carbon black hole“. But not, as he imagines, if the Olympic Dam development goes ahead. No, that massive black hole (at least when expressed in terms of global climate change mitigation) will result from us NOT expanding the mine. Such is the huge energy returned on energy invested (EROEI) of uranium, even when used in today’s ‘inefficient’ once-through thermal reactors. In a future dominated by fast spectrum reactors with a closed fuel cycle, which use vastly more of uranium’s energy content, the above EROEI and emissions equivalence figures just get ridiculous.

Let’s get sensible about nuclear power and carbon emissions, shall we?

Other estimates of life-cycle emissions from world’s best practice are considerably lower than the 60 tonnes CO2-e per GWhe cited in the BHLD study above. For instance, there is this low-end estimate of is 3.3 tonnes CO2-e per GWhe given here:

There is world-wide concern over the prospect of Global Warming primarily caused by the emission of Carbon Dioxide gas (CO2) from the burning of fossil fuels. Although the processes of running a Nuclear Power plant generates no CO2, some CO2 emissions arise from the construction of the plant, the mining of the Uranium, the enrichment of the Uranium, its conversion into Nuclear Fuel, its final disposal and the final plant decommissioning. The amount of CO2 generated by these secondary processes primarily depends on the method used to enrich the Uranium (the gaseous diffusion enrichment process uses about 50 times more electricity than the gaseous centrifuge method) and the source of electricity used for the enrichment process. It has been the subject of some controversy. To estimate the total CO2 emissions from Nuclear Power we take the work of the Swedish Energy Utility, Vattenfall, which produces electricity via Nuclear, Hydro, Coal, Gas, Solar Cell, Peat and Wind energy sources and has produced credited Environment Product Declarations for all these processes.

Vattenfall finds that averaged over the entire lifecycle of their Nuclear Plant including Uranium mining, milling, enrichment, plant construction, operating, decommissioning and waste disposal, the total amount CO2 emitted per KW-Hr of electricity produced is 3.3 grams per KW-Hr of produced power. Vattenfall measures its CO2 output from Natural Gas to be 400 grams per KW-Hr and from coal to be 700 grams per KW-Hr. Thus nuclear power generated by Vattenfall, which may constitute World’s best practice, emits less than one hundredth the CO2 of Fossil-Fuel based generation. In fact Vattenfall finds its Nuclear Plants to emit less CO2 than any of its other energy production mechanisms including Hydro, Wind, Solar and Biomass although all of these processes emit much less than fossil fuel generation of electricity.

Add to FacebookAdd to NewsvineAdd to DiggAdd to Del.icio.usAdd to StumbleuponAdd to RedditAdd to BlinklistAdd to Ma.gnoliaAdd to TechnoratiAdd to Furl


By Barry Brook

Barry Brook is an ARC Laureate Fellow and Chair of Environmental Sustainability at the University of Tasmania. He researches global change, ecology and energy.

27 replies on “Carbon footprint of the Olympic Dam uranium mine expansion”

A couple of years ago BHP said they needed an extra 690MW to expand OD. That may have included thermal needs for the 120 ML/day desal plant at Pt Lowly near Whyalla. When the mine becomes open cut it may be possible to run more machinery on electricity rather than diesel.

SA Premier Rann has been doing the usual hand waving. At one time hot dry rock geothermal (from uranium bearing granite) was going to power everything but that’s not working out. I’m told Cooper Basin gas has only got a few good years hence the extra gas pipeline to Victoria. The Vic border is where a lot of the 900 MW or so of SA wind originates (with 30% CF) and while Leigh Ck coal has 30 years left it has been likened to flammable soil.
Apparently some 2,000 Adelaide homes took up the solar feed in tariff offer; a better number would have been 200,000 homes. There will also need to be a large desal plant near Adelaide as well as the Whyalla plant required for the OD expansion. Adelaide will also become the first major Australian city to have radio controllers for air conditioners.

Given all of the above I expect the May announcement to be unsatisfactory. I think it is likely to involve burning more gas and coal for a smaller OD expansion which is what nuclear critics wanted.


Probably a simpler and more realistic way to look at the expansion of Olympic dam is to consider how much extra power would be used to extract the extra 15,000 tonnes of uranium.
By value copper is about 2/3 of the value(of expanded not present mine), so if we proportion the X3 new capacity is going to be 700MW then the expanded mine would use 470MW, so uranium would be responsible for 2/3×470=313MW.

Interestingly, SA now has >850MW of wind capacity, producing about 300MW of a total average demand of 1300MW, if this expands at same proportion to new demand after Olympic expansion, (1800MW)would expect wind to account for 24% of that 313MW used by expanded uranium, so actually the additional uranium mining will only use 235MW of carbon emitting electricity.
A good solid base-load demand of 700MW demand will probably allow much more than 24% of electricity to come from wind, because this demand will not peak on hot days like residential customers. For PR purposes BHP will probably directly source some of that power for desalination water or for mine use via purpose built wind farms on EP( a good distance from Adelaide’s demand, but closer to Olympic dam and good wind potential sites, presently stranded. Desalination power can be shed at peak demand without any consequences, so will probably sell wind power at market rates and buy “interpretable power” at lower rates, greatly increasing grid stability.

For this reason the Olympic dam expansion could well lower SA’s carbon footprint by increasing wind beyond 24% of state demand, without accounting for the valid benefits of uranium fueled power that you have outlined.


It is true that uranium is a substitute good for fossil fuels. However, this does not imply that were we to not export the uranium from Olympic Dam, there will be nuclear power plants replaced with fossil fuel plants. We should not ignore another substitute good for primary uranium, which is uranium obtained from downblending highly enriched uranium (HEU) with depleted uranium or natural uranium. The HEU will of course mainly come from decommissioned nuclear weapons. Downblended HEU is actually responsible for a very large portion of reactor requirements – see for example the OECD “Red Book” on uranium production.

This raises an interesting question — given that nuclear weapons states can easily meet their reactor requirements by downblending HEU from decommissioned nuclear weapons, should we be exporting uranium to nuclear weapons states at all?


A breakdown of where Australia’s uranium ends up is given here:

Our biggest buyers are the US, Japan, South Korea, Taiwan and the EU (with China likely to be important in the future as their new fleet of AP-1000s come on line), so it’s a mix of nuclear weapon and non-weapon states.

Downblending is done too and I agree it’s important for supply, but that’s linked to broader initiatives beyond energy supply (of which Australia should be a part — as Rudd has stated — but not via conditional directives).


Barry – thank you for the excellent reference material. Just out of curiosity, has there been any discussion by BHP of the possibility of “eating their own dog food” by building a nuclear power plant or two to consume some of the uranium produced at OD and to provide emission free electricity back to the mine?

It would seem like a logical step for a company that wants to expand the market for uranium and that wants to be a good environmental steward.

Also, I thought that the carbon black hole analogy could be interpreted another way – based on your computations, it looks like an expanded OD could be providing a wonderful service for the world in serving as a virtual carbon storage location for all of the emitted CO2 that its uranium is helping to avoid.

Rod Adams
Publisher, Atomic Insights


Rod, yes, I agree. An AP-1000 or ABWR, for instance, would supply sufficient power for both the OD expansion electricity demand and have enough heat remaining to also fulfill a desalination role.

Although as someone pointed out elsewhere, there would be a particularly satisfying message in having the next coal-fired power station in the La Trobe Valley replaced by a NPP. You could tell the people of Yallorn — time for some clean air with >100 times less radiation!


Co-locating the desal with a nuclear plant seems obvious to me. It would be good site even without the desal since there are ample sea currents at that point of the gulf but not so much further along. Despite that there are two lignite fired plants at the head of the gulf at Port Augusta where the cooling water comes from mangrove swamps. Ironically those coal plants may have to expand to provide energy to the Whyalla desal. A branch of the State’s major gas pipeline passes under the gulf to a propane separation plant in the area but supplies are dwindling. Most but not all the locals want new industry and jobs. I believe summer water temperatures often hit 25C which is plenty for RO so the external reactor cooling may not be needed unless some other method is used.

Local intercity transmission is around 130 kv I believe. Since Olympic Dam/Roxby Downs currently has a stranded diesel power station I understand a 300km corridor will be built from the coast with parallel water and electrical lines.

Thus a nuclear plant has to supply electrical and perhaps thermal input to the desal then pumped water and electricity to the distant mine. A major plus would be sending spare water and electricity to nearby towns and the State capital Adelaide if not the whole East Australian grid. Maybe that means two nuclear plants.

Some more factoids about Whyalla; nearby shipping in the gulf brings in coking coal and takes away steel and liquid propane and in future copper concentrate. The Brits detonated some A-bombs 200km west of there post WW2. There’s talk of a 20 MW solar thermal plant in the area and some revhead wants to drive a rocket car 1000 kph on a nearby salt lake.


On the subject of desalination, the upper Spencer Gulf, including around Port Augusta and Whyalla, is a “low energy ecosystem”. These ecosystems are dominated by mangroves and seagrass, which provide nursery grounds for fish. These areas are not appropriate places to release the brine that is a byproduct of desalination. If Olympic Dam is supplied with water from a desalination plant, then the desalination plant should be further south, where there are stronger ocean currents to disperse the brine more quickly.


A Google Earth screenshot of the proposed site for the desal is in this document

Click to access WCN_Special_August_2008.pdf

If various other businesses crowd into that small peninsula there won’t be room to move. It seems to me that Spencer Gulf will undergo ecological change anyway w.r.t. heat and salinity due to locked in warming.


“The expanded mine will be 48 Mt — that is, an additional 36.5 Mt of CO2-e will be released into the atmosphere each year as a result of the Olympic Dam expansion.”

Barry, where did this figure come from?


I quote from above:

Once the full nuclear life-cycle emissions for LWR are accounted for (includes: mining, milling, enrichment, fuel fabrication, reactor construction and operation, decommissioning and storage of spent fuel), the greenhouse gas intensity of the power generated is 60 kg CO2-e per MWhe (BHLD, pg 172), which is 60 tonnes per GWh [for fast spectrum reactors like the IFR, it would be substantially lower, since we skip the mining, milling and storage steps]. Thus in 2005 the emissions equivalent for Olympic Dam uranium mining was ~11.5 million tonnes (Mt) of CO2-e. The expanded mine will be 48 Mt — that is, an additional 36.5 Mt of CO2-e will be released into the atmosphere each year as a result of the Olympic Dam expansion.


By comparison, the expanded Olympic Dam, with its anticipated output of 19,000 tonnes of U, would yield about 94 GWe of average power supply, or 794,000 GWh/yr.

794,000 GWh x 60t CO2/GWh = 47.64 Mt of CO2


So it is calculated on past ore extraction, rather than the deeper more difficult ores at Olympic Dam. As I pointed out the variables that increas your payback period include factors such as ore concentration, ore hardness and ore depth.

Depth is a critical dimenision in termining energy payback (it is potential energy), and the Roxby expansion includes a proposal for a very deep hole.

It also appears your calc does not include the cost (including CO2 cost) of removing the overburden.

So the figures produce an significant underestimate.


The life cycle analysis from UTS has the same range for energy payback (6 to 14 years) as the LCA from storm-smith
Storm-Smith include analyis of an the energy cliff. Give it a read.

The variables that increas your payback period include factors such as ore concentration, ore hardness and ore depth.

Depth is a critical dimenision in termining energy payback (it is potential energy), and the Roxby expansion includes a proposal for a very deep hole. The economics of the proposal change if we were to remove the diesal subsidee given to BHP to truck the overburden and ore out of the hole.

have you calcluated the cost (including CO2 cost) of the overburden portion?

Have you decerned the hardness of the ore?


Storm-Smith’s figures are plain wrong on so many levels — thoroughly debunked here:

To cite the second link (scroll down to the “Energy Lifecycle of Nuclear Power” section):
“It is worth noting that the widely quoted paper by Jan Willem Storm van Leeuwen and Philip Smith (SLS), which gives a rather pessimistic assessment of the Energy Lifecycle of Nuclear Power, assumes a far larger energy cost to construct and decommission a Nuclear Power plant (240 Peta-Joules versus 8 Peta-Joules(PJ)). The difference is that Vattenfall actually measured their energy inputs whereas Willem Storm van Leeuwen and Smith employed various theoretical relationships between dollar costs and energy consumed. This paper also grossly over-estimates the energy cost of mining low-grade Ores and also that the efficiency of extraction of Uranium from reserves would fall dramatically at ore concentrations below 0.05%. Employing their calculations predicts that the energy cost of extracting the Olympic Dam mine’s yearly production of 4600 tonnes of Uranium would require energy equivalent to almost 2 one-GigaWatt power plants running for a full year (2 GigaWat-years). You can follow this calculation here ( This is larger than the entire electricity production of South Australia and an order of magnitude more than the measured energy inputs.”


Only one of the links takes me to a critique of Storm vL. They highlight what prima face appears a bad error. I’ve written to SvL highlighting the critique.

However that cite didn’t address the energy cliff. They quote a favoured LCA of nuclear power showing a payback time of 10 months to 2.4 year that is 6 times more favourable to nuclear than the UTS study that you recommend (5.6 yrs to 14.1 years). Who’s debunking the debunkers? I think this needs further investigation.

BHP first estimated 450MW electical demand, then they estimated the need for 700MW electrical demand (will this be the last rise). Add to this the diesel to truck out 1 million tonnes a day for four years of over burden (each tonne requiring more energy than the last.) Than you can start trucking out the ore and begin count production output (each tonne continuing to require more energy than the last).

What will this average in GWattyears? Maybe Storm van Leuwen were closer to the mark than given credit for?


Correction: I’m not yet confident of the The 1 million tonnes per day figure, (sourced from anonymous leak).

I’ve seen a different estimate of 1 million litres of diesal per day.

Hopefully this can be clarified by BHPs disclousure.


More thoughts after dabbling with Google Earth. Instead of locating the desal to supply OD at Whyalla it could co-located with a reactor on the Great Australian Bight. From Ceduna for example a water pipeline to Roxby might be just 7Okm or so longer. Sea currents in the Bight must have tremendous cooling capacity. After all the next land mass is Antarctica. Build the long awaited trans-Nullarbor HVDC line with some spare capacity to also carry wind, x-othermal, WA peak gas etc.

Specs would be
1) one or two Gen 3 reactors in the Ceduna area next to a desal
2) water pipeline and AC transmission to Roxby Downs
3) a national East-West multi gigawatt HVDC line.

It would work out cheaper than Rudd’s fibre optic to every home proposal.


But will you be able to down-load movies in 10 seconds? After all this is much more important that replacing coal fired electricity, haven’t you seen all the demonstrations in major cities calling for an end to 3 minute downloads, bring on super high speed! All those school computers will go to waste if the kids cant watch cartoons and video games.

Just look at the uptake of high speed(3minute) broadband. ( I am on 12kb/sec unwired only good for reading text and still photographs)


Here in SW Tas we are supposed to be getting 512kbps satellite broadband costing $3k per house. Paid for by Kevin. I’ve had my name down for 6 months.

A truly national grid with nuclear and other desert coast inputs far away from the leafy suburbs has got to have some good points. Kev should go for wireless and satellite rather than fibre optic and save most of the $45bn for a low carbon grid.


[…] Have you been able to pin down the reasoning for Lenzen’s ISA analysis coming up with the CO2 figure of 60g/kWh for nuclear, when none of the studies you cite in Table 1 (or that I’ve seen elsewhere) come close to that. I suspect it involves placing too much weight on SLS-type ‘analyses’ and too great a focus on low-grade U ores – but I’d certainly like to get to the bottom of it. The issue was explored somewhat in the comments of this thread. […]


Thanks for another excellent post, Barry.

You’ll often see anti-nuclear activists bring up Olympic Dam as an example of energy input and resultant life-cycle carbon dioxide emissions from nuclear power. Of course, they cherry-pick Olympic Dam because it certainly isn’t a typical uranium mine.

Since Olympic Dam is basically a large copper mine (and integrated copper smelter, which uses a significant amount of energy) which also produces a little bit of uranium and gold on the side, the energy inputs for Olympic Dam, per tonne of uranium oxide produced, are far higher than any other uranium mine anywhere in the world.

The energy inputs to typical uranium-only mines – Beverley, Ranger, or the overseas mines, take your pick – are tiny by comparison.

Despite this, however, the total energy input to Olympic Dam, including all the energy input for the smelting of the huge amount of Cu produced, is only a tiny fraction of the clean energy output from the mine in the form of uranium – even if we assume that that uranium is used inefficiently in LWRs. If the uranium is used efficiently in fast reactors, the EROEI for Olympic Dam – or any uranium mine – is far, far greater than it is already.

Incidentally, some anti-nuclear activists propose that we should stop uranium mining at Olympic Dam, but keep mining copper – I seem to recall that Jim Green mentioned this briefly on that radio interview that Brook and Green did recently. Of course, that’s completely impractical since the different elements are homogeneously mixed together in the Olympic Dam ore – you can’t mine one without mining the other. Even if you processed the copper but didn’t refine the uranium, you’d still be left with a pile of crude uranium oxide, so you might as well purify and sell this valuable resource.

There used to be a good webpage from BHP Billiton where they listed all the details of CO2 emissions, energy inputs, and other quantitative environmental data for Olympic Dam which can be used to quantitatively show just how big the EROEI factor is, but I cannot find it now.


Thanks Luke, that’s a good point about the energy intensity of OD being substantially higher than other Uranium mines because of the polymetallic nature of the ore body. Using OD statistics for energy inputs is certainly being highly conservative (when trying to point out that Uranium has a low carbon footprint), so antis can hardly complain about this choice.


Luke & Barry,

I would be intersted in the web page you are referring to.
I made a submission on the ODX focussing on the greenhouse assessment, basically because BHPB had made a point in the EIS about covering life cycle emissions of their products and services including scope 3 emissions, yet I found that there was missing data and many exclusions that were not even acknowledged.

BHPB claimed a scope 3 benefit of the uranium saving more emissions than Australia produces (no details on whether this was based on life cycle data), and in making such a claim it is only fair and reasonable that their expansion greenhouse numbers also cover the direct and indirect construction, operating, and closure emissions of the mine and associated infrastructure for roads, rail, water desal, water pumping, power plant, power transmission, ports, the airport and the town.

I did not believe that this was completed and what was done was virtually impossible to disaggregate and check.

Regardless of any support that I may or may not have for any development whether it be a wind farm or a mixed ores mine, a complete data set is required for decisions and comparisons to be made.


Leave a Reply (Markdown is enabled)

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s