The Mactaquac dam has been slowly destroying itself since the 1960s. Alkali-aggregate reaction, the chemistry problem where reactive aggregate and cement alkalis produce an expanding gel inside the concrete, has been swelling the structure for decades. You can’t patch your way out of it. NB Power has known for years it would have to rebuild or retire the 672-megawatt station on the Saint John River, about 20 miles west of Fredericton.
On July 30 the utility executed a development agreement with a partnership led by FlatironDragados, the Broomfield, Colorado subsidiary of ACS/Dragados, alongside Canada’s Aecon Group and Green Infrastructure Partners. Program cost is estimated at $5.4 billion to $6.4 billion.
Twelve months before anyone decides to build
The agreement doesn’t commit NB Power to construction. It opens a 12-month development phase in which the team refines design, schedule and cost, after which the utility decides whether to proceed. Site preparation begins in summer 2026. Construction is expected to start in the second quarter of 2027, subject to approvals, with completion anticipated in 2039.
That structure is the point. Progressive design-build has spread through North American heavy civil precisely because owners got tired of hard-bidding work with unknowable existing conditions. On a sixty-year-old concrete gravity structure with a known degradation mechanism, nobody can price the scope until they’ve cored it, modeled it and staged it. Fixing a number before that work happens produces either a bid nobody can honor or a contingency nobody wants to pay for.
Why the model is spreading
Owners get a real cost before they commit capital. Contractors get paid to do the investigation instead of eating it during pursuit, and they get to walk if the numbers don’t work. The trade is that the owner carries the risk of spending a year and $50 million to learn the project is unaffordable, which is still cheaper than learning it in year four of construction.
If you chase dam, lock or powerhouse rehabilitation work in North America, this is the delivery structure you’ll be asked to price. The same logic drives early contractor involvement on complex building work, including phased jobs like the Reno-Tahoe concourse replacement where McCarthy’s contract value was set during design rather than after it.
Extending life to 2068
The rehabilitation targets a service life running through 2068, which is roughly a century from original construction. New Brunswick has limited alternatives: Mactaquac is a significant share of provincial generation, and replacing it with anything else means either gas or a very long transmission conversation.
Watch the go/no-go decision in mid-2027. That’s when a $6 billion estimate turns into either a contract or a very expensive study.
Also reported by Aecon.