Dipendra Paneru

Topic

Long-Term Hygrothermal Performance and Climate Resilience of Various Mass Timber Assemblies Across Canadian Climate Zones

Department of Civil Engineering

Date & location

• Thursday, September 10, 2026

• 10:00 A.M.

• Virtual Defence

Reviewers

Supervisory Committee

• Dr. Phalguni Mukhopadhyaya, Department of Civil Engineering, University of Victoria (Supervisor)

• Dr. Lina Zhou, Department of Civil Engineering, UVic (Member)

• Dr. Caterina Valeo, Department of Mechanical Engineering, UVic (Outside Member)

• Dr. Guido Wimmers, School of Construction and the Environment, BCIT (Outside Member) 

External Examiner

• Dr. Miroslava Kavgic, Department of Civil Engineering, University of Ottawa 

Chair of Oral Examination

• Dr. Margo Matwychuk, Department of Anthropology, UVic 

Abstract

Mass timber is being adopted across Canada as a sustainable alternative to construction materials such as concrete and steel, driven by its lower embodied carbon, carbon sequestration capacity, and superior strength-to-weight ratio. However, the long-term durability of mass timber depends on moisture management, as wood's hygroscopic and anisotropic nature creates risks of dimensional instability, mold development and delamination. These risks vary across building assemblies, as walls, floors, and roofs each face distinct moisture exposures governed by their position, orientation, and boundary conditions. This dissertation develops quantitative approaches to assess the hygrothermal performance of these assemblies through physics-based numerical simulation, field monitoring, and machine learning-based prediction, across diverse Canadian climate zones. Chapter 1 provides a comprehensive review of hygrothermal behavior of mass timber with a focus on monitoring techniques and assessment approaches. The first part covers the governing physics of heat, air, and moisture transfer in wood-based assemblies, outlining the evolution of hygrothermal analysis from early steady-state methods to modern coupled simulation tools. The second part synthesizes monitoring techniques across three settings: field monitoring at mass timber buildings; controlled laboratory characterization of hygrothermal properties; and remote infrared thermography including UAV-based platforms. Challenges in sensor calibration, deployment, and data integrity are examined across all reviewed programs. Key research gaps in simulation accuracy, monitoring standardization, and climate change assessment establish the foundation for the hygrothermal investigation of CLT wall, floor, and roof assemblies carried out in the subsequent chapters. 

Chapter 2 investigates how drainage cavity ventilation performs under future climate conditions, a critical design question as Canadian climates warm and moisture loads intensify. Four non-load bearing CLT wall configurations were simulated across five Canadian climate zones under historical and projected warming scenarios of 1.5°C and 3.5°C, with ventilation rate ranging from 2 to 200 ACH. To quantify hygrothermal performance, a first-order drying time constant was introduced, and mold growth risk was assessed using the VTT mold growth index. Results indicate that higher ventilation rates improve drying but remain insufficient to prevent moisture accumulation in humid coastal cities, while cold and dry climates showed inherent mold resistance regardless of ventilation rate. Moreover, air cavities and adhesive layers are found to significantly influence moisture retention, with air cavities paradoxically increasing mold risk under prolonged humid conditions. These findings confirm that assembly configuration must be adapted to climate zones to ensure durable mass timber construction. Chapter 3 investigates the long-term hygrothermal performance of CLT-concrete composite floor assemblies at Tallwood House in Vancouver through an integrated framework combining field monitoring, calibrated 1D and 2D numerical simulations, and three machine learning architectures under indoor-only and outdoor-inclusive feature configurations. Field monitoring confirmed that CLT panels maintained stable moisture throughout the post-occupancy period, demonstrating that moisture management strategies applied from manufacturing through installation were effective under restricted drying conditions. Numerical simulations reveal that calibrated WUFI models, while capturing seasonal drying trends, consistently under or overpredicted field-observed moisture content. Results indicate polynomial regression is insufficient for nonlinear moisture dynamics, while SVR regression with outdoor climatic variables achieved R²=0.95 and ANN trained on field data achieved prediction errors not exceeding 0.08% across all monitored floors, confirming that outdoor temperature, solar radiation, and relative humidity remain significant drivers of moisture behavior even in fully encapsulated CLT floor assemblies. 

Roof assemblies face the most severe moisture challenge of any building component, where horizontal orientation, impermeable membranes, and waterproofing imperfections combine to create moisture accumulation risk. Chapter 4 investigates the hygrothermal performance of 3 ply and 5-ply compact flat CLT roof assemblies across eight Canadian cities under historical and projected warming scenarios of GW+1.5°C, GW+2.5°C, and GW+3.5°C, with four waterproofing membrane leakage rates and an ANN metamodel trained on 256 simulation outputs. Results indicate that rain leakage is the dominant driver of moisture risk, with even 0.5% leakage causing biological threshold exceedance in wet coastal cities, while 5-ply assemblies were more vulnerable than 3-ply due to greater hygroscopic mass and restricted inward drying. Climate warming reduced moisture risk in dry continental cities but provided no mitigation in wet coastal climates. The ANN metamodel performed reliably as a computationally efficient surrogate model. Six months of indoor air quality monitoring at the occupied NCIL building confirmed that formaldehyde and VOC concentrations were governed by room enclosure rather than building-wide ventilation. The overall results across the four chapters demonstrate that the integration of physics-based simulation, long-term field monitoring, and machine learning offers a reliable and scalable framework for hygrothermal durability assessment of mass timber buildings, with direct applicability to climate-adaptive design practice and building code development in Canada.