Flammability Metrics for New Low-GWP Refrigerants

Summary

In response to worldwide initiatives to phase down working fluids that have a high global warming potential if accidentally released, the refrigeration and air-conditioning industry is searching for fluids to replace those that are currently in use.  Alternatives exist, but they are mildly flammable and their application has been delayed due to the lack of codes and standards for their safe use. To meet the simultaneous requirements for good thermodynamic and fluid mechanic performance, low toxicity, low-GWP, and low flammability, industry is going to use blends of compounds.  Since the number of components blended may be up to five, the number of permutations is very large.  To find the optimum blends, existing methods are used to analytically predict the other properties; however, no predictive tool exists to rank the flammability.  Burning velocity is used by industry as a metric for flammability.  The present project will develop analytical tools for predicting the burning velocity of refrigerant blends so that industry can use them to optimize the properties of these blends including flammability.

Description

Objective: To develop the capability to measure and predict the burning velocity of mixtures of low-Global Warming Potential (GWP) refrigerants to allow the ranking of the mixtures in terms of their flammability, to assist in the selection and implementation of the best replacements for high-GWP hydrofluorocarbon (HFC) refrigerants.

What is the new technical idea?
The existing high-GWP HFC refrigerants will be phased down, and new low-GWP refrigerants are needed.  Unfortunately, the new low-GWP refrigerants are likely to be mildly flammable, and this characteristic represents a new risk that industry must now consider.  To optimize the thermodynamic and thermophysical properties of the refrigerants, while minimizing the flammability, industry will use blends of the new compounds.  NIST has developed (and industry is using) experimental data and predictive methods for the other parameters, but not for flammability.  Hence, the capability to predict the flammability of pure compounds as well as arbitrary blends would be very useful.   The laminar burning velocity is the preferred flammability metric upon which standards are being developed.  The laminar burning velocity of select blends of compounds will be measured and predicted numerically.   Modeling of the burning velocity will provide an understanding of the physical properties of the compounds that influence their behavior in both test methods and in full-scale fires, facilitating their safe use.

What is the research plan?
The research plan encompasses three tasks.  In Task 1, the possible approaches for measuring burning velocity of low-GWP, mildly flammable refrigerants will be evaluated.  There is a wide range of experimental methods available for measuring the burning velocity; however, the measurements are more challenging for the present compounds because their burning velocities are very low; i.e., near the flammability limit (in the range of 1 cm/s to 8 cm/s, as compared to 37 cm/s for typical hydrocarbon-air systems).  Several methods may be suitable, and each has advantages, disadvantages, uncertainties, and experimental complexities.  It would be very helpful to the HVAC industry to have guidance as to the best way to measure the burning velocity for its needs.  Hence, the first task in the present proposal is to do an assessment of the experimental methods for burning velocity which may be useful as a standard test for marginally flammable refrigerants.  We have used many of the experimental approaches for measuring burning velocity in the past, and two possible experiments are operational now in our lab.  Part of this task will be to participate in the standards meetings (e.g., ISO/TC86/Panel 8; ASHRAE 34, Flammability Subcommittee) to understand the needs and concerns of the relevant parties and provide technical guidance.  Also, since the burning velocities of the fluids of interest are much lower than for the typically-measured hydrocarbons, innovative approaches for data collection and analysis will be explored and developed.

As mentioned above, industry plans to use blends of compounds to achieve the desired physical, thermodynamic, and flammability properties.  Given the time and cost of a single burning velocity measurement, and the very large number of mixtures and conditions of interest, measurements of all the burning velocities for the range of blends of interest is not practical.  Hence, some method of interpolation (or slight extrapolation) of the experimental results is required.   Prediction of the burning velocity for pure refrigerant-air mixtures is possible using numerical flame models, coupled with detailed kinetic models of the flame chemistry (which typically have hundreds of species and thousands of reactions).  Task 2 will develop predictive tools for the burning velocity of pure refrigerants and blends.

As a first test system for which to develop the new capability, NIST will measure and simulate the burning velocity of a popular binary refrigerant blend (R32 with R125; i.e., CH2F2 with C2HF5).  Experiments and modeling will be performed for R32 with increasing levels of R125, over a range of fuel-air ratios.  This study is of value to the HVAC industry since this blend - at a fixed component ratio - is the working fluid in nearly all residential heat pumps and air conditioners in the US (R125 is added to R32 to make it less flammable), but the burning properties have not been fundamentally studied. In this simpler but representative system, NIST will 1.) test the predictive accuracy, 2.) develop methods for database generation or rapid calculation of the burning velocity, and 3.) explore the flammability properties with regard to varying stoichiometry and humidity levels. 

The insight gained in this first case will then be applied for extending the techniques to other HFC-air blends, such as those containing R32, R134a, R152a, and R125, as well as blends of those refrigerants with pure hydrocarbons (e.g., C3H8 and C4H10).  This effort will serve as validation of the kinetic models for the combustion of the pure HFCs, a test of the predictive approach for blends, as well as an assessment of the utility of the experiments for characterizing the burning velocity. 

In Task 3, NIST will extend the measurement and predictive tools developed above using the existing HFC refrigerants to the new low-GWP refrigerants (e.g., R-1234yf, R-1234ze(E)), and their blends, with and without added water vapor.  These substances will include compounds and blends identified as being of interest by the Air-Conditioning, Heating, and Refrigeration Institute’s (AHRI) Low-GWP Alternative Refrigerants Evaluation Program (AREP).  The outcome of the project will be technical input to the standard test method being developed by industry, and predictive approaches and tools for characterizing the burning velocity (and hence, flammability risk) for new low-GWP refrigerant blends to be used by industry.