And that echoes the broader issue underpinning the radical change Labour and the Greens are planning for the electricity sector as a whole.
There are two ways of answering the question.
The one preferred by the smelter's owners, and Invercargill Mayor Tim Shadbolt, yields the answer "almost nothing".
It is based on production costs.
The Manapouri power station's "fuel" - Fiordland's rainfall and the altitude of Lake Manapouri - are provided gratis by Mother Nature, they argue, and its capital costs were incurred back in the 1960s (apart from a second tailrace tunnel added in the late 1990s).
"We have paid for Manapouri several times over. Meridian makes a fat profit on the power it sells us. We are the ones losing money here. Mate, be reasonable."
That is the gist of Pacific Aluminium's case.
Meridian's response is that it has already offered the smelter a substantial discount, worth hundred of millions of dollars on the price freely negotiated between them in 2007.
And in the recent dry conditions, since March 1 Manapouri has only generated enough electricity to supply 30 per cent of Tiwai Point's consumption, it says. The balance has come from other plants in New Zealand.
But there is another way of thinking about what it costs Meridian to supply the smelter, and it is the way a board of directors and most economists would go for.
It is the opportunity cost, or what Meridian could earn by making Manapouri's output available to the rest of the country instead.
Figuring out what that would be is a task of nightmarish complexity, given the scale of Manapouri's output relative to national demand.
It would depend on how the other generator/retailers responded.
But whatever the answer, it is lot more than almost nothing.
The fact that there is an opportunity cost to Meridian is recognised in the contract which Pacific Aluminium is seeking to renegotiate. The exact terms are a deep, dark commercial secret, but a certain amount is known about them.
The power price formula is some function of average wholesale electricity prices, world prices for aluminium, and an adjustment for inflation.
We also know, thanks to some recent prime ministerial indiscretion, that the first "break point" in the contract is in 2016 and that if Pacific Aluminium wants to close the smelter it would be phased out over the following 2 years.
Before the Labour/Greens announcement, the chances of a deal between Meridian and Pacific Aluminium did not look particularly good.
Neither company's owners are committed to the smelter's on power long-term future, at least not to the extent of a subsidising it - from profitable parts of its business in Rio's case, or through accepting a smaller dividend stream in the Government's case.
There is a gap between two price paths, and the further out you look, the wider it gets.
One path is the maximum price it makes sense for the smelter's owners to pay, above which it is worth more to them dead than alive.
The other is the minimum price below which Meridian would be better off making the electricity it currently commits to the smelter available to the rest of the country and taking its chances about coming out a winner in the subsequent industry adjustment. It has generating assets and a retail business to consider.
The basic arithmetic is that the smelter consumes around 14 per cent of the country's electricity output - roughly 5000 gigawatt hours (GWh) a year.
Put another way, that represents about 10 years' normal growth in national demand for electricity.
In recent years demand growth has been weak, while generation projects committed to before the recession have come on line, creating enough overcapacity to weigh on prices, or to provide a window of opportunity for structural change.
If Manapouri's output were made available to the rest of the country, a rational adjustment would take the form of retiring or mothballing generation capacity which thereby became surplus to the system's requirements.
It would accelerate the shutdown ("putting into long term storage" is the polite term) of Genesis Energy's four old coal-fired units at Huntly, a process already under way. That would remove about 2000GWh of annual output or 40 per cent of the overhang.
Attention would then turn to the three large combined cycle gas turbine power stations, Contact's Taranaki and Otahuhu B plants and Genesis' e3p at Huntly. It is a case of "three into two won't go".
Whether commercial terms can be negotiated among the generators to smooth such a transition or whether they would slug it out in the market is one of the many uncertainties complicating the modelling of future price paths.
But that exercise is based on the existing market model, which Labour and the Greens intend to scrap.
Under their model a new Crown agency, New Zealand Power, would act as a single buyer sitting between generators on one side, and electricity retailers and large industrial users on the other.
It would buy power from each generator at a price set not by the market but on a regulated basis, covering the generator's operating costs and providing a "fair" return on its assets.
The assets would be valued at historic cost, probably adjusted for inflation, rather than the current book value which reflects revaluations based on anticipated future earnings under the market model.
NZ Power would sell electricity to the demand side of the industry on terms which reflect the average, rather than the marginal, cost of generation.
Labour's finance spokesman David Parker argues that this level playing field will foster competition among electricity retailers.
To keep the lights on, NZ Power would forecast how much new generation will be required and by when, then invite tenders to supply it. The successful bidder would know what price it would get for that energy, improving bankability.
The cost of additional generation would increase the average cost of power at the wholesale level, but the average would still be dominated by historic costs.
This is a very different system from the current one, where the market is designed to see prices climb to the long-run marginal cost - the level at which the next cheapest generation option becomes viable.
In terms of the Tiwai negotiations, market reform means Pacific Aluminium would be entitled to tell Meridian something like this: "you guys are no longer in a position to do a long-term deal. When there is a change of government you will only be able to sell wholesale electricity to this new outfit, and we will have to negotiate a contract with it. And it will be at prices grounded in the cost of producing power, not opportunity cost derived from a market liable to gaming and designed to drive wholesale prices rise to the long-run marginal cost. That's more like it."
It could make a short- to medium-term deal easier to reach.
The Opposition's move undermines Meridian's negotiating position in a more subtle way as well, by politicising the electricity sector.
By all accounts the Government has scrupulously respected Meridian's operational autonomy under the state owned enterprises model, staying out the negotiations until the point when Meridian told it that it was about to inform the market an agreement was unlikely.
The Government then made the Australians an offer intended to partially the bridge the commercial gap in the short term, but made it clear it was not interested in subsidising the smelter long term.
When the offer was rejected it shrugged and has been happy to let the negotiations revert to Meridian. People who have dealt with its chief executive, Mark Binns, in previous roles rate him a formidably tough negotiator.
That hands-off approach, coupled with public indifference, if not hostility, towards the smelter - outside Southland, at least - has created an environment for the negotiations somewhat different, perhaps, to what the smelter's owners are used to at the state level in Australia.
As a state owned enterprise Meridian's board and management have a statutory obligation to run it as if it were a private business.
The prospect of a shift from that to dealing with a brand new Crown agency, set up not to make a profit but to deliver lower prices for consumers, might well look attractive from Pacific Aluminium's point of view.
Electricity price not the only threat
Aluminium smelters tend to get built to provide a use for large but remote energy resources.
That might be coal in western China, natural gas in the Persian Gulf or Siberian hydro.
In the late 1960s when the Tiwai Point smelter and the Manapouri power station were being built - each to serve the other - they fell into that category.
But Manapouri is no longer remote.
Patrick Strange, chief executive of national grid operator Transpower, estimates that it would take three summers and cost around $100 million to do the work down south to enable all of Manapouri's power to flow north instead of south. Transpower's upgrade to the link between the two islands, called Pole 3, is due to be up and running next month.
The other fundamental threat to Tiwai's future is that Rio Tinto, the Anglo-Australian mining giant which is the ultimate owner of 79 per cent of it (the rest belongs to Japanese company Sumitomo) has been badly burned by a collapse in aluminium prices and is reducing its exposure to the metal worldwide.
It concluded its current contract with Meridian Energy in 2007 at a time when it was very bullish about aluminium and spending US$38 billion to acquire the Canadian aluminium major Alcan - just in time for a global financial crisis and recession.
Last year it wrote US$11 billion off the value of its aluminium business, including Pacific Aluminium, the portfolio of Australasian assets it has hived off for divestment.
As well as Tiwai, Pacific Aluminium includes Boyne Smelters and the associated to Bluff smelter, as owner looks for a way out.
Gladstone Power Station in Queensland, the Bell Bay smelter in Tasmania, the Tomago smelter in New South Wales and the Gove alumina refinery in Northern Territory.
It's a substantial business. But in Rio's financial statements, Pacific Aluminium is not included in the aluminium section of its segmental accounts; it is down the bottom, among "other".
Amid depressed aluminium prices and global overcapacity, a trade sale of Pacific Aluminium looks unlikely any time soon.
An alternative way Rio could get Pacific Aluminium off its books would be a demerger, where shares in it would be allocated to Rio's shareholders and the company separately listed on the ASX.
As for the outlook, Oleg Deripaska, chief executive of Russian producer Rusal, told The Australian this month that 25 to 30 per cent of global aluminium production is at or below the break-even point. "We need to reduce output by 10 to 12 per cent to be able to have a situation in three years where demand will justify the price," he said.
