Foundations cast — Joseph K. Otieno's first month on site

My name is Joseph K. Otieno, MICE. I joined DEMAR as Site Engineer for The Dagaz in Q1 2025. This is a note from the first month at Raptor Road, written in the tone I keep my notebook in — short, dated, and answerable.

The first morning on a site is always quieter than the brochure suggests. Hoarding, gate, security cabin, two skips, a temporary water connection. The architectural model lives in the sales gallery on Lower Kabete Road, two kilometres away. The first month is about what does not yet exist: levels, lines, and the discipline that will hold every later pour to its tolerance.

We started with the survey. The site was set out against three independent control points and re-checked by a second surveyor on day three. Lars-Henrik Mortensen, our principal architect, walked the lines on day five — not because the survey needed an architect's blessing, but because the architect needed to feel the proportions in the dust. Some decisions, he said afterwards, sit better once you have stood at the corner where the lobby will be.

The pile cap programme was the substantive work of the month. Twenty-eight caps, sequenced in pours of four. The mix design is a 40 N/mm² blend specified to keep embodied carbon as low as the structural envelope allows. Each truck on each pour was sampled in pairs of cubes, logged under chain-of-custody, and dispatched to the laboratory for seven-day and twenty-eight-day tests. The seven-day results came back inside specification on twenty-six of twenty-eight caps. The two outliers — caps eleven and twenty-two — were within the published tolerance but flagged for re-test against the twenty-eight-day result. The twenty-eight-day result for both came back clean. Filed.

Cap two is the one we rebuilt. The published surface tolerance for honeycombing on a structural pile cap is five millimetres on the visible face. Cap two, on first strip, presented a small honeycomb at four millimetres on the south-east edge. Within tolerance. We rejected it. The trade rebuilt the section, the quantity surveyor — Knight Mwangi Quantity Surveyors, who are the independent QS on the project — countersigned the rectification, and we moved on.

People sometimes ask why a site engineer rebuilds work that is within tolerance. The answer is that a tolerance is not a target. A tolerance is the worst we will accept. If we accept the worst, the next pour that drifts will drift past it. Holding above tolerance early is how the building stays at tolerance late.

Bowman Adair Kenya, our independent audit firm, walked the site at the end of week three. Their first quarterly construction audit is in their workplan. They asked for the cube test register, the rectification log, the daily site diary, and the sub-contractor induction records. We had them in one folder. That was the point of the first three weeks: to set up the project record so it could be read by anyone, in any order, without surprise.

The crane goes up next month. The construction phone-line is live; any neighbour can call the published number and expect to be answered within the working day. Erik Søndergaard came to site at the end of the month, stood at the corner where the lobby will be, and asked the same question the architect had asked. We are answering it the same way: with the work.

The slow part is over. The visible part begins.