Basement Fire Computer Modeling
Demonstrates how modeling can help predict fire growth and spread within a variety of residential basement scenarios that all feature unprotected wood ceilings.
WHY BASEMENT FIRE MODELING MATTERS
Basement fires are an extremely dangerous challenge for firefighters. In the late 1970s, fire deaths inside a structure occurred at a rate of 1.8 deaths per 100,000 structure fires. By the late 1990s, the mortality rate had risen to three per 100,000.1 Fire Engineering notes that a “large majority of firefighter fatalities or significant injuries occur at what were ultimately basement fires.”2 For these reasons, it is critical to understand the particular safety risks associated with basement fires through experimentation and advanced engineering analysis. To build on and further the knowledge gained from physical experiments, UL relies on computational fluid dynamics (CFD)-based fire modeling tools to expand the available experimental dataset and deepen insights.3
Today, the combination of larger homes, open floor plans, synthetic fuel loads and new construction materials speed up the stages of fire development, creating more hazardous fire conditions.4 One important contributor to the increased fire safety risks – meriting close examination – is engineered wood products, which are increasingly used to help meet the demand for environmentally sustainable and economical building products. Under fire conditions, lightweight engineered floors can lead to greater risk of structural failure in a shorter time period as a result of the reduced cross-sectional dimensions of engineered products relative to traditional lumber floors. When there is a basement fire, or a fire that started in the basement, once on the scene, the fire service is often unclear regarding how long the fire has been burning, the type of floor system exposed to fire conditions and the structural stability of the floor system.
In our previous “New Dynamics of Basement Fires” article, we detailed the results and insights from a variety of research experiments UL conducted to better understand the response of residential flooring systems to fire.5 In addition to those research experiments, we generated both thermo-mechanical finite element models to simulate the structural response 6 and CFD-based fire models to simulate the fire dynamics and effects of different ventilation schemes. In this article, we describe the effects of different ventilation schemes on basement fire dynamics through our fire modeling work.
WHAT DID UL DO?
Using data from live experiments of a basement fire with an engineered wood I-beam ceiling, we built a CFD-based fire model of the same experiment, using Fire Dynamics Simulator software, and compared the results. The model for a fully ventilated condition generally replicated data from the experiment, providing first-order validation. As the fire dynamics depend strongly upon ventilation conditions,
we ran several different models, rather than conducting more expensive large-scale fire experiments, to determine this sensitivity. The other models included changing the ventilation conditions by opening or closing a door or window at different points in time. The choice of scenarios aimed to help examine the quality of the model predictions over a range of ventilation regimes, from fully ventilated to under-ventilated.
Based on modeling the different scenarios, we found that:
- For the Fully Ventilated Basement scenario, where all doors and windows were open for the entirety of the simulation, model predictions compared quite well with experimental data. As compared to an experiment, where data is only available at a few discrete points, the model of the basement fire generated temperatures (gas and thermocouple), airflow, smoke movement and other variables over the entire domain of the basement. This allowed for greater insight into the dynamics of the fire. The model captured some of the key features, such as the movement of air and heat along the channels created by the engineered I-beam structure of the ceiling, leading to heat and flames at the doorway to the first floor.
- With some of the other scenarios where a door or window is suddenly opened after a fire has started, the model showed how the timing of the door’s or window’s opening and the location of the opening could have a dramatic effect on the fire dynamics.
- Using modeling in advance of experiments can potentially provide insights into better and more useful placement of sensors within the structure.
This work was part of a 2010 DHS grant UL received. Our purpose was to help assess and advance the state of CFD modeling of compartment fires. In this research, compartment fires represent a residential basement with an unprotected engineered wood ceiling with a variety of openings.
UL continues to support and advance the use of fire modeling tools for practical and important topics, such as the effect of different ventilation strategies on firefighting outcomes. Extending experimental and field data with innovative modeling techniques is one way UL is helping advance safety science.