Variability of radon levels in Finnish energy-efficient schools: Distribution models and indoor air quality factors

Abstract

This study aimed to investigate indoor radon concentrations in energy-efficient educational buildings, assess the influence of selected environmental and building-related factors, and identify the behaviour of the most suitable statistical distribution model for describing radon data in naturally low-radon conditions. This research has examined the variability of radon concentration levels in 3 energy-efficient school buildings in Espoo, Finland. Within them, 15 Radon detection sensors with a lowest range value (LRV) of 7 Bq/m3, besides 15 Indoor air quality sensors, have been used in 15 classrooms, and the outcome has been analysed to various extents. The results have indicated that a censoring approach based on the sensors’ lowest range value is the best practice for handling radon data and that radon data follows a lognormal distribution, which fits better than other models and shows better consistency when censoring approaches are used. The lognormal model was best fitted in all classrooms on all floors; however, the Akaike information criterion (AIC) of the model fitting was higher on the second floor of the evaluated buildings, where the contribution of subsoil and building material overlapped more. Also, the model was best fitted on the third floor of School 3, where the effect of building material became dominant, implying that modelling on higher floors has the potential to be more straightforward than those closer to the ground; however, more research needs to verify this outcome. It was observed that radon levels were found to decrease progressively with elevation in floor level; however, this trend was more pronounced in buildings with higher radon baselines. All three schools met the Finnish indoor radon requirement of the design phase and action threshold. The minimum value was zero in all three schools. The maximum concentrations were 38 Bq/m³ in School 1 (second floor), 162 Bq/m³ in School 2 (ground floor), and 25 Bq/m³ in School 3 (second floor), all recorded when ventilation was off or operating at partial power. Thus, the results confirm that Finland has sufficiently employed the radon mitigation methods. Radon accumulation was higher in classrooms with no windows or connection to the outdoors, indicating that the potential for natural ventilation might be more effective than the floor in low radon-baseline environments. Radon levels were at their peak around 6 AM in all schools, and adjusting the airflow based on the worst-case scenario in the block, used in School 3, was found to be the most effective strategy in controlling radon levels in indoor spaces; however, switching off ventilation out of the occupancy hours for energy efficiency purposes leads to the highest radon accumulation among 3 evaluated schools. Concrete heavy buildings showed more radon accumulation, and buildings in the same geographical area presented a clear synchrony pattern with each other, increasing the determination of synchrony matrix as the temporal window became finer. Radon accumulation showed positive correlation with Relative humidity (RH%), indoor-outdoor temperature difference (∆T), and absolute barometric pressure difference |ΔP|, with correlation factors of 0.11, 0.14, and 0.07, respectively. RH% was found to have a more consistent correlation with indoor radon concentration (IRC) across all spaces; however, ∆T with the highest correlation factor showed some inconsistency and |ΔP|in airtight and energy-efficient buildings, was proved not to be significant in the whole dataset, and the most inconsistent factor in classroom-specific analyses due to the pressure stabilizing effect of the mechanical ventilation systems.