The Department of Veterans Affairs has tried twice before to replace its aging financial management IT system. Both attempts failed. But the third try looks like it could be the charm.

The current project, called iFAMS, has run into schedule delays and multi-billion dollar increases in its projected lifecycle costs. But as of now, it’s mostly on track, and with billions of dollars at stake in the new system, lawmakers say they’re determined to make sure it stays that way.

iFAMS crossed a significant milestone this month. After five previous small deployments over the past three years, the new system went live in its largest wave yet — increasing its total user base by 60 percent in one fell swoop, to 4,700 users.

Terry Riffel, VA’s deputy assistant secretary for financial management business transformation, said the system is also stable, with uptime of 99.9%.

“This was also the first time VA went live simultaneously with both the finance and acquisition components of iFAMS, which demonstrates this is a viable solution capable of becoming the next-generation financial and acquisition solution for VA,” she told the House Veterans Affairs Committee last week. “It’s important to understand iFAMS is not just a new core accounting and acquisition system — it is crucial to transforming VA’s business processes and capabilities.”

And moving VA off of its legacy Financial Management System (FMS) is a matter of some urgency. The department itself says the old system presents “enormous risk” because it’s becoming increasingly expensive to maintain. And because it wasn’t built to current audit standards, VA has a very difficult time implementing financial improvement recommendations when its auditors make them.

And even though VA is currently able to earn clean audit opinions, doing so requires a lot of manual workarounds, said Nick Dahl, VA’s deputy assistant inspector general for audits and evaluations. The material weaknesses auditors are reporting in VA’s financial statements tie directly back to the inadequacies of the current IT system.

“If VA doesn’t succeed with iFAMS, continuing material weaknesses would be likely, and they’d still be doing the manual workarounds for certain things,” he said. “I would say bigger picture, you have to look at the fact that VA recognized the need to replace FMS more than two decades ago,” he said. “If this system is not successful, how much longer will it take to get a modern system? How much effort, how much financial resources? We dealing with a system that’s 40 years old, and it’s not meeting the needs. So it really is vital that VA does all they can to get this system online.”

And even though VA is confident in the system as of now, it still has a long way to go.

4,700 users is a big population for most organizations, but the total size of the VA workforce who will eventually need to use the system is about 125,000 people. And so far, iFAMS has only been tested in the National Cemetery Administration, parts of the Veterans Benefits Administration, and some smaller VA components. So far, the department hasn’t tried to implement it in its largest and arguably most complex component: the Veterans Health Administration.

“I’ll acknowledge that VHA is the largest organization that we have yet to implement, but I would also tell you that that’s by design,” she said. “We want to make sure that we’re addressing any improvements that we need to make with our deployment strategy before we tackle VHA and also the complex programs that VHA has remaining. Those obviously would impact veterans in some way if we don’t do them correctly. And so we’ve purposely established the schedule in the manner that we have so that we can ensure that we’re learning — and by the time we get to VHA, we’ll leverage all of those improvement activities.”

But VA’s Congressional overseers are concerned that that the department has spent a significant amount of money already to serve a relatively small population of users. VA has put about $1 billion into iFAMS so far, and its total lifecycle cost estimate has grown to about $7.5 billion — up dramatically from an earlier cost estimate of $2.5 billion.

“From the information we have, the system seems to be relatively successful in those offices where it’s been implemented, but there’s still reason to be concerned,” said Rep. Matt Rosendale (R-Mont.), the chairman of the House VA technology modernization subcommittee. “These organizations only add up to a few thousand users and a small fraction of the VA’s budget. Implementing the reforms in the major organizations like the Veterans Health Administration and the big spenders within the Veterans Benefits Administration keeps getting pushed out, and now is not scheduled for a rollout until 2024 and beyond.”

And Riffel said there’s no way to guarantee the program won’t undergo more cost growth between now and the late 2040s, when iFAMS is expected to reach the end of its useful life.

“There will be instances from time to time where we find, for example, a new interface that wasn’t originally identified. We’re doing constant modernization across the enterprise, so we are going to have discovery from time to time,” she said. “What I would tell you is that the way that we’re structured in an agile fashion has allowed us to continue to move other waves forward while we’re getting more intel on a modernization interface or something like that. So although you are seeing some increases, what you’re also seeing is our ability to flex with that and to ebb and flow as our modernization efforts continue.”

One major complication VA will face as it starts to deploy iFAMS into the Veterans Health Administration: the system will need to interface with other huge IT systems that haven’t been built or finished yet, and are facing challenges much more significant than the ones iFAMS has seen to date.

The biggest examples are VA’s future system to manage its medical supply chain, which VA restarted last year after $2.6 billion in spending on a previous, failed effort, and its new electronic health record, whose rollout is currently paused because of ongoing implementation problems at VA health facilities.

“We’ve been working with VHA for the past two years, but we’re at the point now where we need to start really working in earnest on the implementation, and we’re going to start by integrating with their legacy systems,” Riffel said. “As they continue to move out [on modernized systems], we’ll be prepared for that integration. It allows me to integrate with what’s available right now at a site, and also understand that when the enterprise supply chain solution becomes available, we’ll pivot and integrate with that future solution. We know we need to do it, but in the interim we’re going to integrate with what’s available right now so we can proceed.”

Rosendale said Capitol Hill’s view is that failure isn’t an option, and particularly in light of VA’s recent history of multibillion dollar stumbles involving large enterprise IT systems, Congress will be involved in oversight early and often.

“The financial management business transformation program has to succeed, and despite some significant bumps in the road, I do believe that it can,” he said. “We need to see the iFAMS system rolled out to the key offices that handle the majority of the VA budget. This committee will be watching to see if the improvements that have been promised are actually delivered, and we will be monitoring whether this project is able to hit its timelines and recover its delays. I’m committed to keeping it moving forward. No more projects can be permitted to fail again and again, yet continue to receive funding from the taxpayers. That is unacceptable.”