Compute the design snow load on a roof using four major codes: the Japanese Building Standards Law, Eurocode 1 (EN 1991-1-3), ASCE 7-22, and Canadian NBCC. Change the city, slope, shape, thermal condition, and exposure to see the uniform load S_r, total load, drift, and equivalent depth update in real time.
Parameters
Code
Sets the basic coefficient C_b of the design equation
Roof section showing the uniform snow layer, the leeward drift accumulation, and a slide-off arrow when the slope is steep. Color intensity scales with S_r.
kg/m² to kN/m² conversion (g = 9.81) and equivalent depth using ρ = 300 kg/m³ (intermediate settled snow).
Roof Snow Load — BCA, Eurocode 1, ASCE 7
🙋
I always hear that "shoveling roof snow is hard work in snow country", but does structural design actually treat snow weight as a real number? I'd guess Tokyo and Niigata are wildly different.
🎓
Yes, snow is treated as a fully quantified design load. In Japan, Article 86 of the enforcement order of the Building Standards Law defines a ground snow load and "heavy snow zone" map by city. Tokyo is about 30 kg/m², Sapporo 200, and Joetsu in Niigata can reach 350–400 kg/m². You multiply that S_g by shape, thermal, and exposure factors to get the unit load S_r on the roof. ASCE 7 and Eurocode 1 use almost the same form — only the coefficient values differ slightly.
🙋
Switching the code selector on the left from "Japan" to "Eurocode 1", "ASCE 7", "NBCC" changes S_r noticeably. For the same Tokyo S_g, I'm seeing 0.7× to 1.0× — what causes that?
🎓
Good catch — that's the basic coefficient C_b. Japan's Building Standards Law assumes the ground snow basically all lands on the roof, so C_b = 1.0. ASCE 7 assumes "wind blows some of the ground snow off the roof", so the standard value is C_b = 0.7. Eurocode 1 and NBCC sit in between at 0.8. So for the same city and the same roof, an American code design ends up about 30% lighter — and applying the Japanese code to a US building leads to a conservative result.
🙋
It's interesting that S_r drops as I increase the slope. The "roof load vs slope" chart hits zero at 60°. So if I make a steep triangular roof, I don't have to worry about snow load?
🎓
In theory, yes. Above 60° the snow slides off and the roof-surface load is essentially zero. But the snow that slides off ends up concentrated as a drift — on lower roofs below, on eaves as snow cornices, or upstream of obstacles like PV panels. This tool shows the drift as 2× the uniform S_r. Most documented collapses (such as arcade roofs in Tokamachi 1981 or supermarket roofs in Russia 2004) were "uniform load was OK, but drift concentration overloaded a local member".
🙋
OK, so when I'm designing a long-span warehouse in a heavy-snow area, what do I check first?
🎓
The order is: (1) don't get S_g wrong — always check the local heavy-snow designation; (2) pick the roof shape and slope to set μ(α); (3) identify every step and adjacent building where drift can form; (4) set C_t to match reality. Unheated warehouses must use C_t = 1.1; using C_t = 1.0 like a heated house underestimates by 10%. Then check S_r × 1.5 as the "100-year design value" for the right column, so unusually heavy seasons still have margin.
🙋
I noticed you're converting snow density at 300 kg/m³. Fresh snow vs spring granular snow is supposed to be very different — how is that handled in design?
🎓
The design S_g is already given as weight per unit area, so density is only used here to convert back to an equivalent depth for visualization. In northern Japan, March snow is mostly granular at 400–500 kg/m³, and once spring rain soaks in it behaves like 600 kg/m³ "water-loaded snow". Most historical collapses follow this pattern. So in snow country you should also check the "snowmelt + rain" combination, not just the peak snow weight.
Frequently Asked Questions
The basic equation is S_r = C_b·C_e·C_t·I_s·μ(α)·S_g. S_g is the ground snow load (kN/m²) from regional data, C_b is the basic coefficient (0.7 to 1.0 depending on the code), C_e is the exposure factor (exposed 0.8, normal 1.0, sheltered 1.2), C_t is the thermal factor (heated 1.0, unheated 1.1, greenhouse 0.85), I_s is the importance factor, and μ(α) is the slope reduction. This tool evaluates the equation simultaneously for BCA, Eurocode 1, ASCE 7-22, and NBCC.
Up to 15° there is no reduction (μ=1.0). Between 15° and 30° the load decreases linearly to about 0.5 at 30°, and from 30° to 60° it decreases further. Above 60° the snow slides off completely. However, asphalt shingles have more friction than metal, so the reduction starts later. Even steep roofs must be checked for eaves drift and snow shedding onto adjacent surfaces.
A drift is wind-driven snow that piles up on one side of the roof, concentrating on the lower portion of stepped roofs, on the leeward side of taller adjacent buildings, and at re-entrant corners. This tool displays a simplified drift value equal to twice the uniform S_r. In real design, ASCE 7 and Eurocode 1 use detailed formulas that can give 2 to 3 times the uniform load. Overlooked drift loads are the most common cause of roof collapses.
Not at all. Snow density ranges from 50-100 kg/m³ for fresh snow, 250-350 kg/m³ for settled snow, and 400-500 kg/m³ for granular spring snow — a 5 to 10× range. Design loads must assume the worst combination during the season; in northern Japan, March granular snow plus spring rain often produces the maximum load. This tool uses an intermediate density of 300 kg/m³ to back-calculate equivalent depth.
Real-World Applications
Residential and small-building design: Wood-frame houses and prefabricated garages have their roof trusses and beams sized from the local ground snow load S_g. Even at the same floor plan, Tokyo's 30 kg/m² and Sapporo's 200 kg/m² lead to very different beam depths. In Niigata's heaviest snow zones, designers can take credit for periodic roof shoveling. Switching cities and slopes in this tool makes the 5–10× load difference visible immediately.
Long-span warehouses, factories, and logistics centers: Unheated warehouses use C_t = 1.1, and cold-storage facilities require additional load. For roof spans over 30 m, drift and stepped-roof checks are mandatory, and the layout of adjacent buildings alone can double the design load. The February 2014 Kanto-Koshin heavy snow event collapsed many long-span roofs in Japan, with insufficient drift evaluation cited as a frequent cause.
PV mounting structures and agricultural greenhouses: Greenhouses can use C_t = 0.85 (assumed to melt snow), but you must also consider power-outage scenarios. PV mounts are usually installed at 10–20°, where the slope reduction is minimal, so S_r ≈ S_g — making snow load the governing case for arrays in northern Japan and the snow belt.
International project comparisons: North American projects use ASCE 7-22, European projects use Eurocode 1 with national annexes, and Canadian projects use NBCC 2020. The basic coefficient C_b alone causes a 0.7× to 1.0× spread for the same climate, so designers often cross-check multiple codes and adopt the conservative result. This tool's code selector enables instant sensitivity studies during early design.
Common Misconceptions and Pitfalls
The biggest trap is assuming that the roof is safe once the uniform load S_r is acceptable. Most documented collapses are caused by drift concentrations, so the local maximum, not the average, governs. This tool uses a simple 2× factor for drift, but stepped roofs and adjacent buildings can produce 2–3× and locally up to 5× using detailed code formulas. Identify every step, re-entrant corner, and adjacent height during plan review and explicitly list the drift locations.
Next, the misconception that a steep roof is automatically safe. The roof-surface S_r does drop with slope, but the snow that slides off creates new problems: (1) it concentrates on lower roofs, (2) it forms cornices at the eaves that can fall on people below, and (3) it piles up upstream of obstacles like PV panels. Metal roofs in particular can cause injuries or block walkways when shedding. Steep roofs above 30° need snow-stop fittings and a layout that keeps pedestrian paths out of the slide trajectory.
Finally, do not assume that using S_g at 1.0× is automatically safe. The code S_g is roughly a 50-year return value, and rare 100-year winters can exceed it. The "100-year design value" in this tool is a simple S_r × 1.5 estimate; important facilities (shelters, hospitals, data centers) often use a 200-year return period at 1.8–2.0×, and snow-country public buildings additionally check the historical record snow depth at the nearest weather station. The 2014 Kanto-Koshin event collapsed many structures because actual snow exceeded the assumed values. A 1.5 safety factor is only the minimum insurance — designing for "the unexpected" is the iron rule of heavy-snow design.
How to Use
Enter the ground snow load S_g in kN/m² from your local building code map (Japan: 0.5–3.0 kN/m², Eurocode: 0.4–4.0 kN/m² depending on altitude and climate zone)
Input roof slope in degrees (0–60°); steeper slopes reduce design load via the slope coefficient C_s (typically 0.8–1.0 for shallow roofs, 0.0 for slopes >60°)
Specify roof plan area in m² to calculate total load in tons; the simulator applies thermal and exposure coefficients per selected code (Japanese BSL, Eurocode 1, ASCE 7, or CSA S136)
Review computed roof snow load S_r, equivalent depth (cm), drift load at eaves (0.5×S_r minimum), and 100-year recurrence value
Worked Example
Office building in Niigata Prefecture with pitched roof: ground snow load S_g = 2.5 kN/m², roof slope 20°, plan area 800 m². Japanese BSL applies C_s = 0.95 (shallow pitch) and C_t = 1.0 (standard thermal). Roof snow load S_r = 2.5 × 0.95 × 1.0 = 2.38 kN/m². Equivalent depth = 2.38/0.98 = 2.43 m (24 cm). Total roof load = 2.38 × 800 = 1,904 kN (194 ton). Drift load at eaves = 1.19 kN/m². 100-year design = 3.2 kN/m² (using recurrence factor 1.35).
Practical Notes
Alpine and coastal regions in Europe (Eurocode EN 1991-1-3) require separate altitude adjustments; above 800 m, multiply S_g by 1.04 per 100 m elevation
Unheated structures (warehouses, parking garages) use C_t = 1.2; heated buildings typically C_t = 1.0; wind-exposed ridges may apply C_e = 0.8
Roof shapes (flat, gabled, mono-slope, barrel vault) modify drift coefficients: concentrated eave drifts reach 1.5×S_r; use results as minimum, not maximum, for conservative design
Verify local jurisdiction adoption of Eurocode 1, ASCE 7-22, or JIS A 1450 before final design; snow maps update every 10 years in most regions