Validation of Three Methods of Selecting Moisture Reference Years for Hygrothermal Simulations

Chetan Aggarwal, Maurice Defo, Terrence J. Moore, Michael Lacasse, Sahar Sahyoun, Hua GeXV 2020 · DOI

Abstract

Hygrothermal simulations are necessary to permit analyzing moisture performance when designing building envelopes. Owing to the high computing time and cost of long-term simulations, a common approach is to select representative "Moisture Reference Years" (MRYs) from a long-term series of climate data. This study evaluates three selection methods—Moisture Index (MI), Severity Index (Isev), and Climatic Index (CI)—against full 31-year simulations for brick-clad wood-frame walls in three Canadian cities (Vancouver, Calgary, and Ottawa) under historical and future climate scenarios. The performance was measured by the number of hours the Mould Index (MoI) exceeded a threshold of 3. Results indicate that while no method is perfectly accurate, the Severity Index (Isev) generally outperforms the others in identifying the "worst-case" years for mould growth.

Introduction: The Challenge of Hygrothermal Simulation

Designing durable building envelopes requires understanding how moisture moves through walls. Ideally, engineers would run simulations over decades of climate data to account for long-term weathering and climate change. However, running these high-fidelity simulations for 30 or more years is computationally expensive and time-consuming.

The industry standard is to select a single MRY that represents a severe moisture load. The challenge lies in the mathematical selection of this year; it remains unclear which index best aligns with the actual "worst" year produced by a full-scale simulation. This paper validates these selection methods to ensure they accurately predict years of maximum mould risk.

Methodology

The researchers compared three established methods for selecting MRYs:

  1. Moisture Index (MI): A balance between wetting (rainfall) and drying (vapor pressure difference).
  2. Severity Index (Isev): A regression-based approach predicting damage potential using solar radiation, wind-driven rain, temperature, and humidity.
  3. Climatic Index (CI): Similar to MI but utilizes the Penman equation for potential evaporation to better account for net radiation and wind speed.

Simulations were performed using DELPHIN V5.9 on a standard brick-clad wood-frame wall. The study utilized 31-year datasets for historical (1986–2016) and future projected (2062–2092, +3.5°C scenario) climates across three cities with diverse moisture loads: Vancouver (wet), Calgary (dry), and Ottawa (intermediate).

Key Insights and Ideas

Duration vs. Intensity

A critical finding is that the maximum mould index value can be a misleading metric. A single short-lived peak might not be as damaging as a year where the mould index remains consistently high. Consequently, the researchers used the accumulated hours above MoI > 3 as the primary performance indicator.

Annual variation of mould growth hours and index-selected years for the future climate scenario (2062–2092).

Orientation and Climate Specificity

Hygrothermal behavior is highly sensitive to local conditions and wall orientation. For instance, in Vancouver—the wettest city studied—the North-facing wall showed no risk of mould growth due to low wind-driven rain on that specific orientation. This underscores that climate indices must account for orientation-specific loads.

Key Results

The study found that the Severity Index (Isev) was the most reliable method, correctly identifying the "worst-case" class in 3 out of 4 test cases (excluding Vancouver, where no mould occurred). In contrast, the Moisture Index (MI) and Climatic Index (CI) frequently selected years that simulations revealed to be relatively safe, potentially leading to under-designed building envelopes.

Comparison of simulation-derived worst years versus index-selected Moisture Reference Years across historical (H) and future (F) cases.

In the future climate scenario for Calgary (Cal_F), the Isev method accurately pinpointed 2092 as the worst year, which aligned perfectly with the simulation's findings. The researchers noted that Isev-selected years consistently fell into the same or adjacent 200-hour "class" as the actual worst year, providing a high degree of confidence for engineering applications.

Conclusion and Future Work

While the Severity Index (Isev) showed superior performance, the authors acknowledge limitations: the current Isev application is limited to North-facing walls and brick cladding. Future research will expand this validation to other wall assemblies (e.g., stucco), different geographic regions, and alternative damage functions. Additionally, the team aims to develop methodologies for selecting a series of years to better capture long-term moisture accumulation effects.

Summary updated

Abstract

Hygrothermal simulations are necessary to permit analyzing moisture performance when designing building envelopes. Owing to the high computing time and cost of the long term simulations, a common approach is to select representative year(s), the Moisture Reference Year(s), from a longterm series of climate data. It is assumed that the use of Moisture Reference Year(s) (MRYs) provides equivalent results as those provided using long-term series. The selection of MRY(s) is by itself based on the one of the methods available in the literature. In the present study, three methods of selecting the MRYs were evaluated i.e. the moisture index (MI), severity index (Isev) and climatic index (CI). Simulations were performed using individual years of historical climate data extending from 1986 to 2016 and projected future climate data representing the scenario with a 3.5°C increase in average temperature which is expected to occur from 2062 to 2092. Brick cladding installed on a wood frame wall assembly subjected to the climate of three different Canadian cities was selected for analysis. The cities selected were Vancouver (BC), Calgary (AB) and Ottawa (ON). These cities have differing levels of moisture loads. The year having the mould index value more than 3 for highest number of hours among the individual years was compared with the MRY given by three selected methods. A method was considered to be accurate in terms of the prediction if the year selected by that method gives the number of hours with mould index more than 3 which lies in the same class as that of year having maximum corresponding value. In general, it was observed that none of the methods provides the worst year with 100% accuracy, however for most of the cases, Isev method performs better than other two methods in terms of MRY selection.
BibTeX
@inproceedings{aggarwal2020validation,
  author = {Chetan Aggarwal and Maurice Defo and Terrence J. Moore and Michael Lacasse and Sahar Sahyoun and Hua Ge},
  title = {{Validation of Three Methods of Selecting Moisture Reference Years for Hygrothermal Simulations}},
  booktitle = {XV},
  year = {2020},
  doi = {10.23967/dbmc.2020.146},
}
Figure