{"id":997,"date":"2026-06-25T16:14:54","date_gmt":"2026-06-25T14:14:54","guid":{"rendered":"https:\/\/heatpumpingtechnologies.org\/project64\/?p=997"},"modified":"2026-06-25T16:14:54","modified_gmt":"2026-06-25T14:14:54","slug":"iea-hpt-project-64-webinar-on-the-safe-use-of-flammable-refrigerants","status":"publish","type":"post","link":"https:\/\/heatpumpingtechnologies.org\/project64\/iea-hpt-project-64-webinar-on-the-safe-use-of-flammable-refrigerants\/","title":{"rendered":"IEA HPT Project 64 Webinar on the safe use of flammable refrigerants"},"content":{"rendered":"
\n

Hydrocarbons, and propane (R-290) in particular, are on track to become the standard refrigerant across several heat pump sectors in Europe. The shift is driven by the F-Gas Regulation and REACH, and by the low global warming potential of natural refrigerants. It also brings flammability to the centre of system design and operation. On 25 June 2026, the IEA Heat Pumping Technologies TCP hosted a 90 minute webinar presenting selected results from Project 64, “Safety with Flammable Refrigerants.” <\/strong><\/p>\n<\/div>\n\n

\n

Metkel Yebiyo of the Heat Pump Centre opened the webinar and placed Project 64 within the wider HPT TCP framework, an international cooperation that has operated within the IEA since 1978 and now brings together 20 member countries. Safety in the use of refrigerants sits among the programme’s priority areas for 2023 to 2028. The session drew strong international interest, with around 90 registrations from close to 40 countries and 65 participants attending live. The audience spanned manufacturers, component suppliers, research institutes, and universities, a mix that reflects how directly the transition to flammable refrigerants now touches the whole heat pump value chain. (Download the full presentation here<\/a>)<\/strong><\/p>\n<\/div>\n\n

\n

A coordinated answer to a shared safety challenge<\/strong><\/p>\n<\/div>\n\n

\n

Bj\u00f6rn Palm of KTH Royal Institute of Technology, who leads the project, introduced its objectives and structure. Project 64 aims to support a broader and safer use of flammable refrigerants by improving the understanding of the risks, investigating system designs that keep those risks at acceptable levels, carrying out risk assessments, and feeding recommendations into the relevant standards. The scope centres on heating, cooling, and hot water systems below 50 kW in single family and multifamily homes, with a particular focus on units installed indoors, and it covers both hydrocarbons and synthetic refrigerants. Safety during servicing and at end of life is included.<\/p>\n<\/div>\n\n

\n

The work is organised into six tasks covering technical solutions for limiting risk, the investigation of leak scenarios through CFD simulation and practical testing, leak detection, charge reduction, risk assessment, and dissemination. Participating countries include Sweden, Germany, Korea, France, Austria, and the United States. Palm noted that state of the art reports are planned for each technical task, with further webinars during autumn 2026 and presentations at the Chillventa Congress and the China International Heat Pump Conference, both in October 2026, ahead of the final report. (Download the full presentation here<\/a>)<\/strong><\/p>\n<\/div>\n\n

\n

How propane behaves when it escapes<\/strong><\/p>\n<\/div>\n\n

\n

Jafar Esmaeelian of KTH presented an experimental study of R-290 leakage behaviour. Using a dual cylinder setup combined with infrared thermography, the team measured pressure decay for both gas phase and liquid phase leaks, tracked the cumulative leaked mass, and compared the measurements against theoretical leakage rates derived from the standard incompressible orifice equation and the choked flow equation. For a 0.3 mm leak path, the liquid phase model was found to overpredict the leakage rate, and the results pointed to a clear thermal signature accompanying the release. The findings sharpen the inputs that designers and risk assessors rely on when estimating how quickly a charge can escape. (Download the full presentation here<\/a>)<\/strong><\/p>\n<\/div>\n\n

\n

What the field is actually reporting<\/strong><\/p>\n<\/div>\n\n

\n

Bernd Windholz of the Austrian Institute of Technology presented an Austrian survey on the frequency, location, and detection of leaks in heat pump and cooling systems up to 50 kW. Conducted online between October 2024 and March 2025, the survey drew on tailored questionnaires answered by manufacturers and sales companies, installers, and service technicians, covering the previous three years of field experience in Austria.<\/p>\n<\/div>\n\n

\n

Several findings stood out. Around 47 percent of respondents reported that refrigerant circuit leaks occurred in only 1 to 5 percent of all service incidents, which suggests that leak events are relatively rare but still present. The most common leak locations were Schrader valves, connection points to components, and evaporator heat exchangers, which points design improvements, assembly and quality requirements, and training content towards those specific areas. Windholz also highlighted a divergence between the installed base and new product portfolios that will require different refrigerant qualifications to coexist, and he made the case for safety briefings that reach end customers as well as technicians. The preferred testing practices reported in the survey underline the need for robust, field ready standard methods for leak testing and detection. (Download the full presentation here<\/a>)<\/strong><\/p>\n<\/div>\n\n

\n

Modelling the charge, and the limits of synthetic data<\/strong><\/p>\n<\/div>\n\n

\n

Ma\u00eblle Jounay of EDF, presenting work with colleagues at EDF and Mines Paris-PSL, asked how far a physical model of a heat pump can be trusted to generate the data used for refrigerant charge prediction, a question that matters for the smart heat pumps now being developed for predictive maintenance and early leak detection. Building a charge predictor empirically demands large volumes of experimental data, so the appeal of a validated physical model is that it can generate synthetic training data cheaply across any operating point. To test the idea, the team modelled a 6 kW R32 split system in Dymola and compared a reference version against degraded ones in which key refrigerant storage components, the suction accumulator, condenser filter, and compressor crankcase, were deliberately mischaracterised. The reference model predicted charge to within 1.5 percent, but a single mischaracterised component raised the error to as much as 4.2 percent, with the suction accumulator the most damaging, and two combined errors could compound to 5.5 percent. Jounay’s conclusion was measured: even when the error stays acceptable, the model tends to under or overpredict systematically, and using physical models to generate synthetic data may be more challenging than expected, calling for deeper investigation and careful validation before it replaces experimental data. (Download the full presentation here<\/a>)<\/strong><\/p>\n<\/div>\n\n

\n

Reading the room<\/strong><\/p>\n<\/div>\n\n

\n

Arpit Baranwal of Fraunhofer ISE closed the technical programme with a data driven method for spatial risk assessment of indoor gas dispersion. The approach reconstructs three dimensional concentration fields from a sparse sensor matrix using Kriging interpolation, then applies Monte Carlo simulation to quantify measurement and model uncertainty, reporting flammable volume and safe distance with 95 percent confidence intervals, and uses fault tree analysis to evaluate ignition risk. Tested across five experimental campaigns, ranging from a closed room baseline to internal mixing, external ventilation, a passive chimney, and floor drainage, the work confirmed that R-290 stratifies strongly at floor level because it is heavier than air. External ventilation emerged as the most effective mitigation, with floor drainage offering a viable passive backup for heavy gases, while internal mixing removed stratification at the cost of temporarily enlarging the flammable volume. A central message for practitioners is that risk assessment should rely on the conservative upper bound rather than the mean, since sparse sensing can otherwise underestimate the true size of the gas cloud. (Download the full presentation here<\/a>)<\/strong><\/p>\n<\/div>\n\n

\n

Looking ahead<\/strong><\/p>\n<\/div>\n\n

\n

Together the presentations illustrated the breadth of Project 64, from controlled leak experiments and field surveys to charge modelling and probabilistic risk assessment. The results cannot always be transferred directly between countries, but they offer valuable insight into how and where flammable refrigerant systems fail, and they will help guide technical standards, product design, training programmes, and service procedures as propane becomes more widespread.<\/p>\n<\/div>\n\n

\n

Further outcomes from the project, including the task reports and forthcoming autumn webinars, will be published on the project website. The webinar was recorded and can be accessed here<\/a>, and full project information and outcomes are available at <\/em>https:\/\/heatpumpingtechnologies.org\/project64\/<\/em><\/a><\/p>\n<\/div>\n\n

\n

<\/p>\n<\/div>","protected":false},"excerpt":{"rendered":"

Hydrocarbons, and propane (R-290) in particular, are on track to become the standard refrigerant across several heat pump sectors in…<\/p>\n","protected":false},"author":40,"featured_media":1012,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"_relevanssi_hide_post":"","_relevanssi_hide_content":"","_relevanssi_pin_for_all":"","_relevanssi_pin_keywords":"","_relevanssi_unpin_keywords":"","_relevanssi_related_keywords":"","_relevanssi_related_include_ids":"","_relevanssi_related_exclude_ids":"","_relevanssi_related_no_append":"","_relevanssi_related_not_related":"","_relevanssi_related_posts":"718,983,953,797,967","_relevanssi_noindex_reason":"","footnotes":""},"categories":[50,1],"tags":[],"markets":[],"class_list":["post-997","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-events-activities","category-uncategorized"],"acf":[],"_links":{"self":[{"href":"https:\/\/heatpumpingtechnologies.org\/project64\/wp-json\/wp\/v2\/posts\/997","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/heatpumpingtechnologies.org\/project64\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/heatpumpingtechnologies.org\/project64\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/heatpumpingtechnologies.org\/project64\/wp-json\/wp\/v2\/users\/40"}],"replies":[{"embeddable":true,"href":"https:\/\/heatpumpingtechnologies.org\/project64\/wp-json\/wp\/v2\/comments?post=997"}],"version-history":[{"count":2,"href":"https:\/\/heatpumpingtechnologies.org\/project64\/wp-json\/wp\/v2\/posts\/997\/revisions"}],"predecessor-version":[{"id":1013,"href":"https:\/\/heatpumpingtechnologies.org\/project64\/wp-json\/wp\/v2\/posts\/997\/revisions\/1013"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/heatpumpingtechnologies.org\/project64\/wp-json\/wp\/v2\/media\/1012"}],"wp:attachment":[{"href":"https:\/\/heatpumpingtechnologies.org\/project64\/wp-json\/wp\/v2\/media?parent=997"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/heatpumpingtechnologies.org\/project64\/wp-json\/wp\/v2\/categories?post=997"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/heatpumpingtechnologies.org\/project64\/wp-json\/wp\/v2\/tags?post=997"},{"taxonomy":"markets","embeddable":true,"href":"https:\/\/heatpumpingtechnologies.org\/project64\/wp-json\/wp\/v2\/markets?post=997"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}