SPECIAL ISSUE PAPER The discrepancies in energy balance in furnace testing, a bug or a feature? Wojciech Węgrzyński | Piotr Turkowski | Paweł Roszkowski Fire Research Department, Instytut Techniki Budowlanej (ITB), Warszawa, Poland Correspondence Wojciech Węgrzyński, Fire Research Department, Instytut Techniki Budowlanej (ITB), 00‐611 Warszawa, Poland. Email: w.wegrzynski@itb.pl Summary The paper aims to explain the differences found in the heat release rate measure- ments in a large sample of standard fire tests (EN 1363‐1). A total of 379 tests of ver- tical assemblies was investigated, all performed in furnace SPARK of the ITB Fire Testing Laboratory, in 2015‐2018. The assemblies were subdivided into two groups —wall assemblies and fire‐rated doors. These assemblies were also compared with the results of the test of a wall built with aerated autoclaved concrete blocks that was considered as the benchmark test. It was observed that walls built with highly insulated sandwich panels require less heat to maintain standard thermal exposure conditions (20%‐30% less) than their counterparts built from gypsum plasterboard or aluminium and fire‐rated glass. In case of doors, it was observed that combustible samples required significantly less heat than the benchmark case (40%‐70% less), which indicates that the combustion of the sample inside of the furnace was an addi- tional, significant source of heat release, that may skew the qualitative assessment of their performance in fire. A more in‐depth discussion of the results is provided, with some ideas on the direction of further developments in fire testing. KEYWORDS ASTM E119, EN 1363‐1, fire resistance, ISO 834, standard fire testing, temperature‐time curve 1 | INTRODUCTION 1.1 | Fire testing In the 19th century, the world transitioned from wood and stone architecture to modern materials such as concrete, steel, and glass. 1 One of the essential aspects of this transition was the need for the broader provision of fire safety to the occupants—often requested after multiple major fires that engulfed large parts of cities (eg, Great Fire of Pittsburgh of 1845, San Francisco Fires of 1851 and 1906, or The Great Chicago Fire of 1871). With the industrial revolution, the capital invested in factories became more concentrated in space, and more focus was put on protecting it. Besides the aspect of the citi- zen safety and protection of the capital, the resiliency of wooden buildings was disputed. The large conflagrations have shown that many stone buildings survived the fires. However, there was a need to prove this ability in a reliable and repeatable manner. This lead to the birth of ad hoc testing methods of the behaviour of building elements (slabs, columns, walls, and later doors, gates etc), which eventually evolved into fire testing as we know it today. This was possible through a significant simplification of the complex fire phe- nomena into a set of standardised thermal exposure conditions, to imitate the effects of a fire. These effects were described with the use of so‐called temperature‐time curves, out of which the most prominent is the “Standard temperature‐time curve,” referred to as ASTM E119, 2 ISO 834, 3 or EN 1363‐1 4 curve (Figure 1). In the European framework, 4 this exposure is also described in the form of Equation 1. T f ¼ 20 þ 345 log 10 8 t 60 þ 1   ; (1) where T f is the temperature in the furnace (°C), and t is time (seconds). Received: 18 February 2019 Revised: 2 May 2019 Accepted: 21 May 2019 DOI: 10.1002/fam.2735 Fire and Materials. 2019;1–12. © 2019 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/fam 1