Mark HibbsPakistan’s Next Chinese Reactor

Pakistani luminaries met with Chinese luminaries a few months ago, and their handshake will translate into a brand new 1,000-MW power reactor–Kanupp-2–being plunked down into the middle of Pakistan’s mega-metropolis Karachi. Maybe even two reactors. The unofficial announcement with details is here. In April, China Daily confirmed that there is a foreign contract. So it’s a done deal. Right?

Not quite.

Friends at the Pakistan Atomic Energy Commision (PAEC) cautioned this month that this transaction hasn’t yet been formally approved and concluded. “It’s still pending” was how they described it.

The backstory, I subsequently learned, is straightforward. With questions looming about NSG guidelines and intellectual property rights for these 300-MW-loop projects, Pakistan and China may have their reasons for being more circumspect than most about how they conduct their bilateral nuclear commerce. But that plays no role here. In one very essential aspect this Sino-Pakistan reactor deal isn’t any different from any other transaction anywhere else where a nuclear power plant exporter is selling his wares to a foreign client: Money talks, and each side will try to leverage its political assets to gain commercial advantage.

The money

Whether two 1,000-MW PWRs would cost PAEC $9.6 billion as announced or a single unit could be had instead for about $4 billion, which is what I was told a week ago, that’s a lot more than I think some Western observers have generally assumed this project would cost (these are BTW the same pale faces who have already concluded that this sale is a foregone conclusion on the basis of soft Chinese financing). The starting point is this question: Does Pakistan have billions of dollars to throw at a venture like this? That’s dollars or rupees or RMB per installed kilowatt. If you talk to people hovering around the World Bank or the International Monetary Fund, which have devoted a lot of time and effort trying to understand and help fix Pakistan’s electricity supply problems, the answer is well, no, it doesn’t.

Friends in Pakistan point out that Chashma-3 and -4, each rated at 340-MW, cost $750-million each, with Chinese soft money financing about half of that amount. Hence at $4-billion, PAEC wouldn’t be getting any discount for installing more megawatts at Kanupp-2, even if China would provide $2-billion in financing. To the contrary.

So there’s an alternative scheme which has been floated in Islamabad and Beijing: Pakistan can get the reactor for less, but it would have to agree to cut Chinese industry into a build-own-operate (BOO) venture–a business model that Russia and Turkey are committed to trying out at Akkuyu, and that I hear Chinese investors are interested in pursuing with any of a small number of potential future nuclear power plant clients who are short of cash and perhaps nuclear project management expertise.

The grid

If China and Pakistan go that route, the project will certainly be delayed because of Chinese due dilligence. As Pakistani friends explain, Chinese potential owners of a nuclear power plant in Pakistan will make a hardnose calculation of estimated long-term return on investment. That would take into account a host of factors which would play out in any nuclear plant project anywhere in the world: management and engineering quality, infrastructure, regulation, political risk. But also the state of the power grid. And that is a touchy subject in Pakistan.

The existing Karachi Nuclear Power Plant (Kanupp-1) is connected to the grid system operated by the Karachi Electricity Supply Co (KESC). If Kanupp-2 is built, it will also connect to KESC.

In 2009, the World Bank estimated KESC’s grid losses to be a whopping 30%.  A lot of this loss was due to theft. One year before, the company was privatized on the basis of a $1-billion equity investment by shareholders. Now, if you are going to invest in Pakistan’s power sector, you had better have deep pockets, and KESC’s prime shareholder, Abraaj Capital, a Dubai-based private equity firm, apparently has them. KESC says it is now profitable, and foreign investment in Karachi’s grid, most recently this year by the Asian Development Bank, is continuing. Here’s one narrative which compares the situation at KESC to the rest of Pakistan. Why the confidence in KESC, you might ask. The answer in part is that KESC is today the only vertically integrated power company left in Pakistan, selling to a captive market  in an urban agglomeration–the world’s 11th biggest–of about 15-million people. That looks like shooting fish in a barrel. But for how long? And will customers pay their electric bills? According to this account, as much as 35% of KESC’s power is still being stolen, and the company’s restive workforce is yet another source of uncertainty. Given that it might take 20 years to amortize an investment in a nuclear power plant, it can be assumed that if potential Chinese investors are on the horizon, they will want an answer to those questions.

The dams

It’s no secret that the U.S. isn’t too happy about yet another Chinese PWR–this would be number three–slipping under the wire of the NSG guidelines by means of China’s grandfathering claim, after the NSG’s 46 members abetted this state of affairs by awarding India an exception to those rules back in 2008. I and two colleagues at Carnegie two years ago proposed a possible long-term fix for this, but if that general approach has its merits, and we still think it’s worth consideration, it won’t happen overnight.

In the meantime, the U.S. has been mulling prospects for revving up development assistance to Pakistan to generate a lot of hydropower. According to the State Department in August, at issue is the Diamar-Bhasha dam project which would have a price tag of about $12-billion–and generate 4,500-MW-worth of electricity.

Lest you conclude that the U.S. government in its infinite wisdom, and informed by NSG considerations, is subtly connecting the dots here, Pakistani friends insist that it’s not at all a conspiracy by Washington to try to substitute dams for reactors. They point out that the U.S. isn’t the only one helping Pakistan. Saudi Arabia and the EU are also contributing to Pakistan’s hydropower development.

But the dam projects on the upper Indus figure in PAEC’s strategic planning and in its negotiations with China. Some press reports over the last couple of years identified the first ACP-1000 reactor would be Chashma-5. That site is up in the Punjab closer to where all that river water is supposed to generate electricity. Karachi is near the mouth of the Indus, 1,000  kilometers downstream.  The bottom  line is that Pakistan has to figure out where its electricity is going to come from during the next 20 years. Depending on the answer, the next Chinese reactor  might not be built at Karachi at all but instead somewhere else–and there are six candidate sites scattered about Pakistan which PAEC has selected to build it.

To get back to the Sino-Pakistan reactor negotiation, for PAEC the dam-building has its utility. If Pakistan has other potential sources for generating a lot of electricity, the price for ACP-1000 might come down. Already, PAEC has interjected that the price should reflect the fact that, while China will build the ACP-1000 as Fuqing-5 and -6, right now there is no reference plant, entitling Pakistan to a certain risk discount. That logic takes into account that, for the 300-MW units which PAEC has set up at Chashma, Pakistan is highly confident that these reactors will perform according to design specifications because Pakistan has done most of the safety upgrading and site-specific design and project engineering work, including seismic related engineering, while China never continued with the 300-MW plant design after it finished Qinshan-1 back in 1991. In fairness to China, it must be said that China has built two-loop 650-MW units based on this template to more advanced specifications.

There’s also the question of how much Chinese soft money will be available for financing. That, friends in Pakistan say, will depend directly upon how much of the engineering and procurement work for the project is assigned to Chinese contractors, and, indirectly, upon what kind of countertrade opportunities Pakistan can offer China.

At the end of the day Pakistan might tell China this: Especially if France raises IP issues, we might be your only real market outside China for ACP-1000 for a while. In response, rest assured, China would tell Pakistan: But if you want more nuclear energy we are the only option you have.




  1. rwendland (History)

    The World Nuclear Association says the Chinese nuclear regulator in 2011 instructed CNNC and CGNPG to progressively merge their ACP1000 and ACPR1000+ (latest version of CPR1000) designs. Both three-loop designs originally based/inspired on the same French design, but with different nuclear cores. Many CPR1000s have already been built, and the ACPR1000+ incorporates many Generation III features, like double containment to protect against external explosions and airplanes, improved seismic capability to 0.3 g, increased core thermal margins, a core-catcher and improved operation systems.

    So it looks highly plausible Pakistan would get a Generation III (or near) reactor, if the ACP1000 has been picked up the latest ACPR1000+ features. Do you know of any good technical papers on the ACP1000 design they might get?

    NB characterising it as “being plunked down into the middle of Pakistan’s mega-metropolis Karachi” is rather, umm, unbalanced. It’s actually on the coast 25 km west of Karachi, on scrubby land 4 km from the nearest village going by Google Earth. Still way too close to a major conurbation IMO for a large nuclear reactor, but not in the urban area.

  2. Balachandran (History)

    Just as info:
    A cost estimate was made for C-3/4 in April 2008 based on:

    i) US$ exchange rate 63 Rs /$.
    ii) price escalation of 35% over C-2
    iii) A factor of 18% for the impact of Chinese Yuan appreciation against US Dollar since the signing of the C-2 contract in May 2004.

    i) Total cost of project (including in core fuel) =US $ 2038 Million
    ii) Chinese scope of work= US $ 1785 Million
    iii) Total FE component= US $ 1505 Million.

    i) Two Loan agreements signed with the Chinese in early 2010 for US $ 1474 Million.
    ii) The 2009 estimate for C-3/4 = Rs. 190 Billion with a FE component of Rs. 140 Billion

    The 1000 MWe reactors being offered to Pakistan (to be called K-1000) are of Generation II plus and not generation III.

    • rwendland (History)

      The press reports on the ACP1000 bid/sale to Pakistan are claiming it is a Generation III design, presumably originated from CNNC. Though the distinction between II+ and III is not well defined, so it is somewhat a marketing term. It doesn’t (I think) have passive cooling, but nor does the Areva EPR which is accepted as Gen III.

      The cross-sectional diagram of the ACP1000 given at the unofficial site below shows it having a double containment and what looks like a core-catcher, which suggests it has picked up at least some of the latest ACPR1000+ features.

  3. Balachandran (History)

    There are some major differences between Generation II+ and Generation III reactors. For example, the active length of the fuel assembly in generation III is 14 feet. In Generation II+ it is 12 feet.

  4. Tom (History)

    Great post. Just a very basic question – when push comes to shove, will financial terms really be the deciding factor? As I understood it, C-3/4 were supplied at very friendly rates. While it may be in CNNC’s interest to try to squeeze as much as possible out of the project at this stage, will not geopolitics determine the final outcome?

    • mark (History)


      In a nutshell, geopolitics are the basis of this commerce. I would never for that reason question that these Sino-Pak nuclear reactor deals won’t stop happening. What interested me is that on the ground they seem to be governed by the same kind of decision making that obtains everywhere else where nuclear power plant deals are signed and sealed. In this regard, China is just another supplier, and Pakistan is just another customer. There are however–as in every other nuclear energy project arrangement–some unique considerations. One of them in this case are the murky questions about China’s IPR, given that everyone, including the World Nuclear Association’s WNN site, confidently identifies the ACP-1000 reactor as being a derivative of French PWR technology. Whether that will matter?…we’ll see.

  5. rwendland (History)


    This also looks to me to be primarily a commercial deal. CNNC wants a first foreign sale, to start of a series of foreign sales, and Pakistan is probably an easy starting point (NSG aside).

    China has been saying for several years it will replace the French IPR in this reactor design (and the similar ACPR1000), and given the power output has been uprated by 15% (about 1100MWe gross, up from the 955 MWe French design) requiring a redesign of many components, I have no doubt that they have very seriously gone about this. World Nuclear Association’s WNN site also reports CNNC saying “independently-developed ACP1000” and they have “full intellectual property rights over the design”.

    Areva is unlikely to contend this, as that would upset its essential commercial links with China nuclear companies. As China is about 50% of the market for new nuclear, Areva is unlikely to risk a rupture, and there will be many opportunities to negotiate away any conflicts as a side-deal to some future major contract. And anyway, the French 900 MWe design was likewise derived from a Westinghouse design, so the IPR situation is probably pretty arguable.

    NB Regarding the location of Karachi NPP, I’ve just realised that here in the UK Hitachi and the govt are hoping to build 3 new ABWRs at Oldbury, 12km from the suburbs of Bristol, and 21km from central Bristol. So it is hard to criticise Pakistan for locating a new NPP 25km from central Karachi, when we are building closer in the west. Personally I think in both cases this is way too close to a major conurbation.