On permafrost, a solar foundation that works on commissioning day can fail before the array does. Design for end-of-life ground, choose piles from the site investigation, and prove them with load tests.
Ikayuut’s 2,500 solar panels in the Arctic hamlet of Naujaat are expected to run for 30 years. The ground under them will not stay the same for that long. For Solvest Northern Solar Solutions, PRI designed the foundations for the array’s racking and electrical auxiliary structures on adfreeze and bedrock micropiles, with climate change adaptation built into the loading analysis, and confirmed the design with pile load testing. PRI took the same approach for Solvest farther west in the Arctic, at the Town of Inuvik’s 689 kW Midnight Sun Complex system, commissioned on April 23, 2026.
The panels on both projects are standard equipment. The foundations are not, and on permafrost they carry an Arctic solar project’s longest-lived risk.
The method below draws on those two projects and on two more off permafrost. At Solar North in Masset, on Haida Gwaii, PRI carried out the field investigation, preliminary foundation design and quality assurance and quality control testing for Hedgehog Technologies Inc. At Anahim Lake, PRI ran a pre-production pile load testing program, completed the detailed foundation design and supported construction as Foundation Engineer of Record for the 3.8 MW solar facility. Neither site is permafrost, but the sequence of investigating, testing, designing and then checking what was built applies on all four.
• Investigate where the array will stand. Drill and sample at the array location, not at the nearest borehole on record. Ice content, ground temperature and pore water salinity decide how frozen ground carries load and how it will respond as it warms, and regional data rarely captures them at the scale of a solar site. On Solar North, the field investigation came before the preliminary foundation design, so the design responded to measured ground rather than assumed ground.
• Design for the permafrost at the end of design life. Climate warming and construction disturbance both warm the ground. Project ground temperatures forward over the asset’s life and design for the warmest permafrost the foundation will sit in, not the coldest it sits in today. At Naujaat and Inuvik, climate change adaptation was part of the loading analysis from the start.
• Choose the pile from the ground. Where competent bedrock is within reach, a bedrock micropile carries load into rock and takes the permafrost out of the bearing question. Where it is not, an adfreeze pile relies on the bond between frozen ground and the pile shaft, and that bond weakens as the ground warms, which is why the previous step comes first. Ikayuut and Midnight Sun each used both systems.
• Analyze the racking and the electrical structures separately. Racking and electrical auxiliary structures carry different loads and tolerate different amounts of movement. Treat each on its own terms rather than applying one pile design across the site.
• Prove it with a pile load test, and run it early where you can. The designs at Ikayuut and Midnight Sun were confirmed with pile load testing. At Anahim Lake the load testing came before production, so the detailed foundation design was built on tested capacity rather than assumed capacity. In a remote community the test needs equipment and a crew on site, so plan it into the same logistics window as construction.
• Stay on through construction. Design assumptions hold only if what is built matches them. On Solar North, PRI’s scope included quality assurance and quality control testing as well as the foundation design; at Anahim Lake, PRI supported construction as Foundation Engineer of Record. In both cases the team that set the assumptions checked the work against them.
The fair objection is cost. Site-specific drilling, a climate-adjusted analysis and a load test program all take money and time on a project already paying northern freight, and a conservative generic design could skip them. Over-design is not free either: every extra metre of pile and tonne of steel rides the same sealift, barge or winter road. Testing is how a team finds out which side of that line its design sits on.
On permafrost, a foundation design is a forecast, and the pile load test is the only part of that forecast that can be checked on site.Evaluating a cold-climate or off-grid renewable application? Talk to the PRI Team.