Retrofit vs new-build: why unlocking existing UPS beats new BESS
New industrial batteries cost USD 300 to 500 per kWh installed. Existing UPS batteries are already paid for, already installed, and already sitting idle 99 percent of the time. Here is when retrofit wins.
The default industrial energy-storage answer is still "install a new BESS". It is what integrators quote. It is what most funding is sized for. In a lot of cases, it is the wrong first move.
Every industrial site with meaningful load already runs a UPS. That UPS holds enough stored energy to bridge a grid outage. Often 10 to 30 minutes at the site's full protected load. Ninety-nine percent of the time, it sits idle. Waiting for an outage that may or may not come this year.
That is a huge, paid-for, already-permitted stored-energy asset doing nothing.
The unit economics
A new industrial-grade BESS in Singapore or Australia lands at USD 300 to 500 per installed kWh, before financing and grid interconnection costs. For a 2 MWh system, that is USD 600,000 to 1 million in CapEx. Plus a 6 to 18 month lead time on permitting, procurement, install, and commissioning.
Payback depends heavily on tariff structure and market participation. But even at the best sites, with high demand-charge exposure and active FCAS or Reserve AS participation, you are looking at 5 to 8 years to break even on the hardware alone.
Retrofitting a control layer onto an existing UPS is a different order of magnitude. The battery is already installed. The interconnection already exists. The site's backup obligation is already sized around it.
What you add is:
- An edge controller (a few thousand dollars).
- A safety-guarded dispatch envelope (software).
- A market participation contract (via an aggregator in Australia, a licensed retailer in Singapore).
Total added CapEx: five figures, not seven. Time to first revenue: weeks, not quarters.
When retrofit is the right call
Four conditions have to be true for retrofit to make sense.
- The existing UPS has usable spare capacity above the backup floor. Most industrial UPS systems are sized with 20 to 40 percent headroom above the actual protected load, because the specifier did not want to be caught short. That headroom is monetisable. Sites where the UPS is sized exactly to the protected load, with no margin, get less out of retrofit.
- The cell chemistry can handle cycling. Older lead-acid UPS systems cycle poorly. Every discharge shortens their life. Modern lithium UPS systems cycle well. If the UPS is lead-acid and more than 5 years old, the honest advice is usually "wait for the natural replacement cycle, then install lithium and retrofit at that point".
- The site sits behind a meter with real exposure. Peak demand charges in Australia. Contestable tariffs with peak-window pricing in Singapore. Market participation programs in both. Sites on flat tariffs with no market access do not benefit from retrofit.
- The operator will accept a software-defined backup envelope. You define a minimum reserve, say 60 percent state of charge, below which the controller cannot go. Operators who understand this trust it. Operators who do not tend to panic about it. Retrofit works better with the former.
When new-build is the right call
Two cases:
- Greenfield sites with no existing UPS. Nothing to retrofit. Sizing the new BESS to serve both backup and market participation from day one is usually the cleanest approach.
- Sites where retrofit does not pay off and there is a strong case for a purpose-sized BESS. Some large data-centre corridors in Singapore, or NEM-connected industrial loads with kV-scale demand charges, do justify a dedicated BESS on top of the existing UPS. In those cases, retrofit does not compete with new-build. It complements it.
The compounding case for retrofit-first
For most industrial operators reading this, the retrofit case starts from near-zero regret. You are using an asset you already own. You are not signing a 10-year financing deal. You are not giving up site space to new equipment. If the numbers work at your site, you are earning revenue in weeks. If they do not, you paid a few tens of thousands of dollars to find out. Compared to tens of millions to learn the same lesson with a new BESS.
That difference matters more than any spreadsheet comparison of levelised cost of storage. It is why we usually recommend retrofit as the first move. Even for operators who will eventually need new BESS capacity. Prove the operating model on the paid-for asset first. Then size the next investment on real dispatch data from your actual site.
If you want to run the numbers for your site, talk to us. We do a free retrofit assessment on the first 20 sites per market per quarter.