The Obama administration took a lot of heat -- and rightfully so -- for its handling of $535 million in loan guarantees handed to solar maker Solyndra, a high-risk solar company. But it may be making a bet on nuclear power that could dwarf the Solyndra losses and compound losses for consumers in Georgia at the same time.

The Vogtle Electric Generating Plant expansion being built in Georgia, 45.7% owned by Southern (NYSE:SO), is relying on an $8.3 billion loan guarantee from the Department of Energy to move forward with construction. The plant has been delayed and seen cost overruns, and now its application for a guarantee has been extended for the fifth time without final approval, even though Wall Street has no interest in getting anywhere near the project without government support. Could this end up being a another big loser for the government?

Solyndra was baby stuff
You may remember Solyndra, the solar upstart that got a $535 million loan guarantee despite not having a cost-effective solar module. In rapid fashion, the company built a state-of-the-art plant and quickly found out that it couldn't compete with low-cost Chinese solar modules. Solyndra went bankrupt in 2011, and the Department of Energy got back scraps on its original investment.  

The $8.3 billion being negotiated for Vogtle may seem less risky, but nuclear-plant designer Westinghouse has never completed an AP1000 reactor, the design planned for Vogtle. Then there's the fact that Wall Street wants nothing to do with Vogtle, or nuclear energy right now, hence the need for government support. S&P downgraded the outlook on Southern and subsidiary Georgia Power's credit from "stable" to "negative," the same outlook project partners Oglethorpe Power and Municipal Electric Authority have. High risks and a bad history for nuclear projects is giving rating agencies second thoughts about debt coming from Vogtle.  

How could Vogtle become Solyndra?
The fastest way the government could lose billions on Vogtle is if cost overruns hit the project mid-stream and funding dries up, something we've seen happen in the past. Vogtle 1 and 2 were originally estimated to cost $660 million, but costs ballooned to $8.87 billion by the time they were done.

More recently, Progress Energy's Levy County, Fla., nuclear plant ballooned in cost from a range of $4 billion to $6 billion to a final estimate of $24 billion before being canceled. Customers will end up eating $1.5 billion in costs because of what's called constriction work in progress, or CWIP, which allows a utility to charge for a project before it's completed.

The list of projects that have been delayed or cancelled goes on and on. NRG Energy took a $481 million loss after giving up on its South Texas Project in 2011 (no CWIP for NRG). A Calvert Cliffs Nuclear Power Plant addition has been on the table for years, but with a $9.6 billion cost for a reactor and no loan guarantee, it's unlikely to be built. Plants in Minnesota, Illinois, Iowa, Pennsylvania, and other states barely got off the ground before being cancelled, primarily because of cost.

The challenge becomes similar to Solyndra if constriction is started on Vogtle 3 and 4. If costs go up from a current estimate of $15.5 billion (and if history teaches us anything, it's that costs will go up significantly), who pays for the additional cost? Southern is the majority owner, but it's not in a position to cover billions of dollars in overruns. Consumers are paying for the plant already because of CWIP, but will regulators increase what they're paying just to get the plant running? Wall Street has already weighed in, downgrading the equity and debt outlook, based largely on this project and doesn't seem eager to offer more debt without a loan guarantee. At the end of the day, the government can say no now or run the risk of writing a check that could lead to an $8.3 billion loss or more. It's risking throwing good money after bad. 

Consumers will pay either way
One challenge is that for Southern, the downside of Vogtle isn't as big as you might think. Consumers will pay about $2 billion to Georgia Power during construction, because of CWIP charges to consumers before the plant is finished. If the project is canceled, some of the costs may fall on the utility, but recent CWIP decisions in Florida show that consumers aren't getting a refund, and the utility may even get a return on money spent.

So if the project is cancelled now, Georgia Power consumers pay, if it's canceled before completion, the DOE will lose and consumers will pay. The only upside is if the project is completed on-time and on-budget. I think there's little likelihood of that, given the direction nuclear energy is headed in the United States. 

At one time, Solyndra seemed like a great idea to the DOE, and the Vogtle addition did as well. Maybe it's time to revisit how much of a slam-dunk nuclear is, given the history of cost overruns and construction delays that could cost the government billions.