United Launch Alliance (ULA), the Boeing-Lockheed Martin joint venture that builds and launches Atlas 5 and Delta 4 rockets for government customers under the U.S. Air Force Evolved Expendable Launch Vehicle (EELV) program, is dealing with cost growth — driven in large part by a shrinking industrial base — that has left it open to criticism.

The biggest inflationary driver, company officials say, is the end of NASA’s space shuttle program, which once covered a large portion of the overhead costs borne by ULA’s suppliers, particularly in propulsion. These costs are now falling largely onto the EELV program as the suppliers try to keep facilities open in anticipation of a post-shuttle space exploration program that may or may not materialize.

ULA also is launching more satellites these days, including some under contracts that were negotiated in the late 1990s. The prices in those contracts reflect assumptions of high rocket production rates driven by commercial market demand that never panned out.

The EELV program’s steep budget-growth curve notwithstanding, ULA Chief Executive Mike Gass says the company has reduced the government’s launch costs by 25 percent or more compared with earlier-generation rockets, which was a key objective when the program was conceived in the early 1990s. He also says the future launch prices ULA is quoting to its customers will come down considerably once the uncertainty surrounding NASA’s human spaceflight program is resolved.

Gass spoke recently with Space News Editor Warren Ferster and staff writer Peter B. de Selding.

Why should the U.S. government continue to support two families of EELV rockets when one would be sufficient to meet current demand?

The fundamental thing that drives it in all the studies and all the financial analysis is the fact that you already have two families; the nonrecurring cost is sunk, and the fact that the spacecraft have been integrated in some cases to one family or the other so if you switched you’d have to go through some additional costs to move a particular satellite from one launch vehicle to another.

But wouldn’t it save money in the long run?

We have two products, but we have one team supporting both products. So we’re trying to get the value and the economics of that one system but still be able to utilize the investment and the infrastructure that we created for two systems. There is some value in that diversity but if that diversity comes at too great of a cost we could look at downsizing. But all the analysis says it doesn’t pay off for a fairly long time — the numbers have gone out to 2025 to 2030 as the payback period. With today’s budget crisis the challenge is coming up with that upfront investment for that switch, compared to other priorities.

What’s behind the EELV program cost growth?

You have the normal inflationary cost growth but the No. 1 thing going on in our supply base right now is uncertainty. And the uncertainty is on what NASA is going to do for its exploration and commercial crew needs. For the past decades, whether it was the old system or the current EELV, we shared the industrial base with the shuttle program. Every one of our major suppliers worked on the shuttle program. So the shuttle program was absorbing some of the industrial base and the overhead structure. Going forward we all thought there was going to be some follow-on program. Now we don’t know.

The single biggest factor is propulsion because there are not a whole lot of other programs that use that infrastructure as compared to, say, our electronics manufacturing or our ordinance manufacturing; there are other users of that industrial base. If there is no exploration program we’ll rationalize the industrial base, shut facilities down, and we’ll shed the overhead. But you’re sitting there having to keep it open, waiting, and you can’t take action overnight. If the decision is made tomorrow, it’ll take three, four years to go rationalize that capability.

The cost of the Atlas 5’s Russian-built RD-180 main engine has risen as well. How are you managing that contract?

The cost has gone up but not by an unreasonable amount as compared to Russia’s inflation rate or the value of what we’re getting from an engine performance for the price. It’s still a great value. The unit cost also has risen because ULA has slowed the build rate of engines to ensure we have a steady production flow, by not getting too much inventory ahead of the launch needs. Letting a production line go dark is risky to mission success and expensive for the life cycle because of the high restart costs driven by product requalification test costs. We balance the near-term unit costs with the long-term total life cycle costs of the product line.

Why are Atlas 5 prices in the NASA Launch Services 2 contract so high?

The contract required not-to-exceed prices in it. In our proposal, we were very clear that the not-to-exceed prices assumed some really dramatic changes in the Air Force’s buying practices, and if they changed dramatically this is what could happen to the price. I personally sent the letter to NASA Administrator Charles Bolden and all the acquisition people and said, “Okay, this is how we did the proposal but here are the numbers; here’s how it would be if the Air Force basically stayed constant,” and the number was about 50 percent less than the not-to-exceed. If the Air Force implements their planned new efficient buy, it went down even more. If the Air Force and NASA bought together, it went down even more. But right now from a budgeting standpoint people are budgeting worse case to worst case.

What annual launch rates do you anticipate over the next five years?

It’s starting to get up in the 12 on average rate. The 10 to 14 rate is probably what we’re looking at. And that’s before any other of these adjacent market opportunities develop. Commercial crew — that’s additional. We’re also looking at how we can continue to attract customers from the true commercial market and bring in another launch or two a year.

ULA has talked about reducing staff from the current level of about 3,600 to 3,000 to support a minimum of eight launches per year. At 10 to 14 launches per year, would you be adding to rather than reducing the current work force?

Most likely the total would not go much higher. ULA is committed to finding aggressive efficiency savings through our implementation of our one-company initiatives and commitment to continuous improvement. We expect to get more output with the same resources or the same output with fewer resources on a year-by-year basis. The major areas affected by a change in launch rate are launch and production crews. The engineering work force will be augmented with a few people and potentially change skill sets around. The majority of the core infrastructure will not be affected for this rate increase. Launch has high fixed costs, so increased rates can be accommodated with relatively low marginal cost increases.

Sometimes the launches come close together, requiring ULA to employ multiple crews. Does the customer pay extra for that via Enhanced Launch Capability contracts?

Our first goal is to try to do accommodate the surge without any extra funds — do as much as possible by figuring out what efficiencies and opportunities we can create. When there are gaps in resource levels, we negotiate that enhanced launch capability. So in years past most of the time the additional resources are for launch crew augmentation, because the best solution to compressing the schedule is by adding a second shift to bring launch spans down, so the need is for more launch technicians and engineers and this is enabled by the enhanced capability contract. Sometimes the funding comes from multiple customers depending on the cause of the compression of the schedule between the Air Force, the National Reconnaissance Office, NASA and even a commercial customer.

Given the launch rates you’re anticipating, will there be a significant amount of Enhanced Launch Capability funding necessary in the next few years?

I don’t know what you mean by significant. It’s a relatively small amount. It’s not even a percentage of the total cost and the ability to go get that additional funding, and some of it may have already been negotiated; we may have put it into the price.

Are you moving toward a common upper stage for the Atlas 5 and Delta 4?

We’re on a path towards getting that commonality. Right now we’re on the path using an incremental approach. Why do you need a large nonrecurring investment in the short run that has a very long-term payback unless the launch rate changes dramatically? Not that these are real numbers, but if you had $500 million nonrecurring to develop a new common upper stage and cause all the satellite design reintegration actions and the change is going to save you $5 million a launch, you need 100 launches to pay it back. But any time we have an opportunity where we’re going to have to make a design change, it’s always done in a way that promotes commonality.

What about a common variant of the RL10 upper-stage engine?

We’re in the preliminary design phase heading to the critical design review of that common engine. Basically we have an A and a B model; the core C model would fly on both Atlas and Delta. Eventually we’re going to do it; it’s just a matter of when.

Have you ever considered opening ULA’s books so that critics would better understand why the EELV cost figures are what they are?

Anybody in the U.S. government that needs to make a decision has access to every number inside ULA.

How do you view the prospect of competing against Space Exploration Technologies or some other company for Pentagon business?

We formed ULA because our nation came to the conclusion that there wasn’t enough business for two competitors. Our customers couldn’t afford two infrastructures, two teams. We have one team, one infrastructure, but provide the benefit of two fully capable systems. I struggle with what’s changed in the market. If it was the right decision in 2005, when we announced the formation, what’s changed in 2011?

ULA has five unsold Delta 2s. At what point do you throw in the towel and write them off?

You hate to run the line in the sand, but four years ago we thought it was two years ago. Years ago we used to think of the Delta 2 as a standalone team of people; they had to be totally separate. The Delta 2 team is now integrated into the EELV launch crew, and if a Delta 2 launch comes together we go support it. We stop charging against the EELV contract; we charge against the Delta 2 line, complete the launch and then when they’re done, they go to work back on whatever their job was before. So we took away the problem of this standing army waiting around for a launch. This year we have several Delta 2 launches so we have some people dedicated to it, but we don’t need to have that dedication anymore. But right now, there are not a whole lot of satellite opportunities in the Delta 2’s market space. We know of every one that’s out there and there are not many that need that performance.

Peter B. de Selding was the Paris bureau chief for SpaceNews.