← Back to blog

Why Your Corner Apartment Is Colder — and Costs More to Heat

October 5, 2026 · 5 min read

“Why is my apartment cold when my neighbor's identical one isn't?” is one of the most common complaints about Latvian apartment living — common enough that consumer-advice sites like kopaa.lv run standing pages collecting exactly this complaint, and LV portāls has published explainers on it repeatedly over the years. The usual answer, when the two apartments really are the same size and layout, comes down to position in the building: corner and end apartments have two exposed walls facing outside instead of one.

The physics, not just the complaint

This isn't a vague folk theory — it follows directly from the EN ISO 6946 heat-loss methodology this calculator already uses. Heat loss through a wall is proportional to its exposed area: double the exposed wall area at the same construction, and you double the heat lost through it. A corner apartment isn't slightly worse off than an interior one — it's running a second full exterior wall nobody budgeted for.

Take a typical 602 series panel apartment (~50 m², 250mm claydite concrete walls, the most common building type in Latvia). An interior unit has one exposed wall, around 40 m². A corner unit has two, around 80 m². The wall construction — and therefore its U-value — is identical in both cases:

Unit positionExposed wall areaWall U-valueHeat lost through walls, per season
Interior apartment~40 m²~1.6 W/(m²·K)~5,875 kWh
Corner/end apartment~80 m²~1.6 W/(m²·K)~11,750 kWh

Calculated per EN ISO 6946 using Rīga's heating-season climate defaults already built into this calculator (ΔT ≈ 19.9°C, 4,608 heating-season hours). Same wall construction in both rows — only exposed area differs. Floor and window losses are on top of this and are identical for both unit types, so they're left out of this comparison to isolate the wall effect.

That second wall alone is costing a corner-apartment resident roughly €476 over a heating season at this season's district-heating tariff (€0.08096/kWh, Rīgas Siltums, effective 1 October 2026) — about €74 extra per month while the heating is on, before any other difference between the two apartments is even considered.

This season, that extra cost is worse than usual

District heating tariffs across Latvia are up noticeably this season. The national average has risen from roughly €67–75/MWh last season to around €93/MWh now, and several towns have approved increases of 2–23% since September — Rīga's own tariff may rise a further 20–30% from December, driven by natural gas prices that have climbed about 129% since the start of the year.

Source: press.lv, “Тепло — но дорого” (Oksana Petrenko), citing SPRK (the public utilities regulator)

Why “average” building calculators don't catch this

Most free heating calculators available to Latvian apartment owners estimate cost from floor area and a self-reported energy class, averaged across the whole building. That approach can only ever produce one number per building — it has no way to tell a corner unit from an interior one, even though the physical difference between them is real and, per the calculation above, roughly doubles wall heat loss. Only 36.1% of Latvian households made any energy-efficiency improvement in the last five years (Centrālā statistikas pārvalde, 2023) — many are likely still paying this corner-position premium without knowing it has a specific, calculable cause.

Check your own apartment's position

Calculate heat loss for a corner or end apartment

Select the corner-unit preset for your building series — pre-filled with two exposed walls instead of one.

Wall-area and heat-loss figures in this article are calculated using this calculator's own EN ISO 6946 methodology and material data (claydite concrete λ ≈ 0.55 W/(m·K) per EN ISO 10456), not estimated separately. Tariff and market figures are attributed to the sources linked above. This article is for informational purposes only — actual costs depend on your specific building's condition, heating system, and consumption patterns.