Clinic Addition on a 1962 Concrete Frame: Snow Drift or Drift-Free?
A single-storey 1962 reinforced concrete clinic in a northern city (ground snow load 2.1 kPa, specified 1-in-50) is being extended with a 6 m wide, 4 m tall mechanical penthouse placed 1.5 m back from the high-roof edge. The existing roof is nominally flat (slope 1:50) with a parapet 0.9 m high; the penthouse roof is flat. The reviewing engineer notes the original drawings show 125 mm slab spanning 4.5 m with 12 mm bars at 200 mm, and no snow drift was considered in 1962. The architect's sketch shows the penthouse roof draining toward the main roof through two 100 mm scuppers. The owner wants to know whether the existing roof framing near the penthouse needs strengthening before the addition is built.
1. Determine the drift surcharge the penthouse imposes on the main roof and explain how the geometry affects whether a drift case governs.
Read the worked answer
Start with the drift parameters rather than the nominal load. For a 4 m tall, 6 m wide obstruction, the drift height is taken from the standard drift relation using the upwind fetch and the height difference; with a fetch of roughly 30 m and h = 4 m the calculated drift height is on the order of 1.2–1.5 m, but it is capped at h and at the leeward-roof geometry. The drift width is typically 4 times the drift height, so a 1.4 m drift spreads about 5.6 m across the main roof - well beyond the 1.5 m setback, meaning the full triangular surcharge lands on the existing slab.
Key judgement: the 1962 design never included drift, so the existing slab near the penthouse is being asked to carry an additional trapezoidal load of up to about 0.7–0.9 kPa at the peak (0.4 times the drift height times the snow density, roughly 3 kN/m³). Check the slab's factored capacity against 1.25D + 1.5S with the drift included. If the slab is governed by flexure near midspan the drift at the edge may be tolerable; if it is governed by shear or by negative moment at the support, the local surcharge can control. Recommend either relocating the penthouse to reduce the drift or adding local strengthening - do not simply rely on the scuppers, which affect rainwater, not snow accumulation.
2. The architect proposes lowering the penthouse roof to 2.5 m and moving it to the edge of the building. Does that change the drift case?
Read the worked answer
Yes, and in two opposing ways. Lowering the height difference from 4 m to 2.5 m reduces the drift height proportionally (drift height scales roughly with the height difference for a given fetch), so the peak surcharge drops. But moving the penthouse to the edge removes the leeward roof on one side, so the drift can form only on the remaining three sides, and the edge condition may instead create a corner drift or a windward drift against the parapet.
The correct approach is to re-evaluate the drift geometry for each roof step rather than assume the original case still governs. Also check whether the reduced height still qualifies as an obstruction requiring drift - very low steps may fall below the threshold where drift is required. Finally, confirm the parapet at the edge does not itself create a drift pocket against the lowered penthouse roof. The lesson: drift is a geometry problem, not a single number, and moving an obstruction can move the governing load case.