Towards standardized safety protocols for iron-based energy carriers: International alignment through round robin testing on safety characteristics

Aleksandra Semenova; Stefan H. Spitzer Sr; Barbora Dvorakova; Enis Askar; Patrick Funnemann; Roy T.E. Hermanns; Vojtech Jankuj; Alexis Vignes; Richard Kuracina; Zuzana Szabova; Frederik Norman; Dieter Gabel; Enrico Danzi; Ekaterina Makarova; Ivan Lorenzon; Adrian Toman; Wojciech Adamus; Conrad Hessels

doi:10.12688/openreseurope.22839.2

Towards standardized safety protocols for iron-based energy carriers: International alignment through round robin testing on safety characteristics

Aleksandra Semenova ,
Stefan H. Spitzer Sr ,
Barbora Dvorakova
,
Enis Askar
,
Patrick Funnemann ,
Roy T.E. Hermanns ,
Vojtech Jankuj ,
Alexis Vignes ,
Richard Kuracina ,
Zuzana Szabova ,
Frederik Norman
,
Dieter Gabel
,
Enrico Danzi
,
Ekaterina Makarova
,
Ivan Lorenzon
,
Adrian Toman ,
Wojciech Adamus ,
Conrad Hessels

Article

Authors & roles Authors

Peer review record Peer review

Metrics

Cite

Section

Research Article

Published

June 3, 2026

DOI

https://doi.org/10.12688/openreseurope.22839.2

Versions

June 3, 2026 (Published Manuscript Under Review 2.0)
February 18, 2026 (Published Manuscript Under Review 1.0)

Slovak Research and Development Agency

Grant ID(s): APVV-21-0187
European Commission

Grant ID(s): 101186946
TKI-Chemie

Grant ID(s): CHEMIE.PGT2022.005

Views

144

Downloads

Citations

See full metrics

MeCRE, Safety Characteristics, Chemical Energy Carriers, Clean Energy Transition, Standardized Iron Fuel, Dust explosions

Abstract

Background

Iron powder appears to be a promising solution for long-term energy storage and (inter-) continental transport, as it is safe to store and does not require energy to maintain its state, unlike, for instance, liquefied hydrogen. However, while the fundamental research is well underway, large-scale implementation is still in its early stages, with a growing number of promising demonstrators emerging.

Methods

This article contributes to the large-scale implementation of iron as an energy carrier by presenting a round-robin test of four iron powders currently used in research and larger-scale demonstrators. These powders were tested on their safety characteristics in the standard 20 L apparatus across eight European countries.

Results

The resulting data are intended to support future standardization efforts using different iron samples as standardized fuel. All tested powders were classified either as non-explosible or as belonging to the category of marginally explosible dusts (Class 1). This provides a clear picture of the level of explosion protection measures that need to be considered for the safe use of iron powders in energy carrier applications.

Conclusions

Along with that, the study detected variations in the results and pointed to shortcomings in the current standards that may cause such discrepancies. These findings emphasize the importance of improving testing procedures to support standardization and ensure the safe use of iron powder as an energy carrier using an a-priori-approach rather than subsequent testing.

Corresponding author

Stefan H. Spitzer Sr

0000-0002-6851-9963

s.spitzer@eifi-tech.eu

Grant information

This project has received funding from the European Union’s Horizon Europe research and innovation programme under grant agreement No 101186946 (Centre of Excellence for Safety Research (CESAR))). The powders tested were produced during the TKI-Chemie project (Lighthouse MEC II, CHEMIE.PGT2022.005). Part of this research (measured in Slovakia at MTF STU) was supported by the Slovak Research and Development Agency under Contract No. APVV-21-0187.

© 2026 Aleksandra Semenova (Co-Author); Stefan H. Spitzer Sr (Corresponding Author, Co-Author); Barbora Dvorakova (Co-Author); Enis Askar (Co-Author); Patrick Funnemann (Co-Author); Roy T.E. Hermanns (Co-Author); Vojtech Jankuj (Co-Author); Alexis Vignes (Co-Author); Richard Kuracina (Co-Author); Zuzana Szabova (Co-Author); Frederik Norman (Co-Author); Dieter Gabel (Co-Author); Enrico Danzi (Co-Author); Ekaterina Makarova (Co-Author); Ivan Lorenzon (Co-Author); Adrian Toman (Co-Author); Wojciech Adamus (Co-Author); Conrad Hessels (Co-Author)

This work is licensed under a Creative Commons Attribution 4.0 International License.

First version published

February 18, 2026 (https://doi.org/10.12688/openreseurope.22839.1)

Latest version published

June 3, 2026 (https://doi.org/10.12688/openreseurope.22839.2)

Revised

Amendment

The revised manuscript addresses a broad range of methodological, presentational, and scientific concerns raised by the reviewers.
On the methodological side, the authors clarified how explosion parameters are determined in line with EN 14034 and ASTM E1226 standards, and updated tables to include standard deviations where multiple experimental series were available. Graphical representations of explosion overpressure and rate of pressure rise as a function of dust concentration were added, accompanied by a discussion of the observed deviation from typical literature trends, particularly the absence of a distinct pressure maximum for iron dust. The injection dynamics at high dust concentrations are now more explicitly addressed.
Particle size data, previously presented in a potentially misleading box-plot format, have been replaced with a table reporting d10, d50, and d90 values per laboratory along with the measurement method used which is a clear improvement in transparency. The discussion of sources of experimental scatter has been expanded, and concrete recommendations for future studies are now included.
Structural and presentational issues have also been resolved: figure–text alignment was corrected, y-axes now correctly indicate overpressure, K_St values were added to the tables, and the comparison of iron with other energy carriers was moved from the conclusion to the introduction where it belongs. Several phrasing improvements were implemented throughout. In total, ten figures were added to the revised manuscript, and one existing figure was modified, reflecting the extent of the visual revisions made.
Two points merit continued attention: the introduction's literature survey, while revised, should be carefully verified for referencing accuracy; and the absence of EDX analysis on regenerated samples, though acknowledged, remains a limitation the authors attribute to practical constraints rather than scientific reasoning.

1.

Introduction

The need for alternative carbon-free energy sources is becoming increasingly crucial in achieving the goals of the Paris Climate Agreement. Although solar and wind power are promising solutions for energy production, their inherent intermittency in both time and location requires efficient energy storage and transport.

The use of metals as energy carriers to decarbonize the energy sector appears to be a promising solution. This approach offers several key advantages: metals are among the most energy-dense zero-carbon energy carriers, are non-toxic in bulk form, relatively easy to recycle, and allow energy to be stored over long periods without losses.¹^–³

In particular, iron is not classified as a critical or conflict material⁴ and is abundantly available across all continents and countries.⁵ If renewable energy is available, the iron energy carrier cycle can operate in a loop, allowing the stored chemical energy to be released as heat (see Figure 1). Alternatively, it is also possible to react iron powder with water vapor, resulting in the release of hydrogen. Thereby, the iron powder effectively acts as a hydrogen carrier.

Figure 1. Schematic of the cycle using iron as an energy carrier: If renewable energy is available in surplus, the iron oxide gets reduced. If energy or heat is needed, the iron is burned.

Table 1. Comparison of selected energy carriers with respect to energy density and explosion characteristics.

Energy carrier	Energy density (MJ/kg)	MIE (mJ)	$p_{\max}$ (bar g)	K_St/G (bar·m/s)
Iron	7 [11]	n.a.	5.2 [12]	50 [12]
Aluminium powder	31 [13]	3 [14]	12.5 [12]	1100 [12]
Brown coal (dust)	10 [15]	160 [12]	9.1 [12]	143 [12]
Bituminous coal (dust)	24 [16]	60 [17]	9.1 [12]	86 [12]
Wood dust	17 [18]	7 [12]	9.2 [12]	102 [12]
Crude oil	47 [15]	n.a.	n.a.	n.a.
Gasoline	44–46 [15]	0.8 [19]	7.9 [20]	47.66 [21]
Diesel	42–46 [15]	n.a.	7.4 [20]	n.a.
Methane	50–55 [15]	0.29 [20]	7.1 [20]	56.7 [20]
Hydrogen	120–142 [15]	0.017 [20]	7.3 [20]	976.6 [20]

The chemical reactions governing combustion (Equation 1) and hydrogen production (Equation 2) are:

(1)

4 Fe + 3 O_{2} ⇌ 2 {Fe}_{2} O_{3}, Δ_{R} H_{298 K} = - 1640 kJ / mol

(2)

3 Fe + 4 H_{2} O_{(v)} ⇌ {Fe}_{3} O_{4} + 4 H_{2}, Δ_{R} H_{298 K} = - 129 kJ / mol

The re-charging phase with hydrogen reduces the iron oxide formed during combustion back to iron and water (Equation 3):

(3)

{Fe}_{2} O_{3} + 3 H_{2} ⇌ 2 Fe + 3 H_{2} O_{(v)}, Δ_{R} H_{298 K} = + 86 kJ / mol

The concept of using iron for long-term energy storage has been known for more than a century⁶ and is now gaining renewed momentum due to the need for a safe and dense energy carrier to solve the inherent intermittency of renewable energy. Although iron as an energy carrier is largely absent from mainstream discussions,⁷^–⁹ the landscape is rapidly changing. Start-ups such as RIFT, AMBARtec AG, Fenix Energy, and Iron+ are actively rolling out pilot plants and launching the first commercial installations.¹⁰

However, as with many energetic materials, iron powder can pose an explosion hazard when dispersed in air, for example during handling or transport, where it may form explosible dust–air mixtures. Table 1 provides an overview of the key explosion characteristics of iron and other common energy carriers. Compared to many conventional solid, liquid and gaseous energy carriers, iron dust generally exhibits a lower explosion severity.

In order to enable large-scale deployment, it is essential that the safety characteristics of iron dust are well understood, and that appropriate explosion protection measures are formally addressed through standardised guidelines. This paper makes an important contribution by establishing a solid foundation for such efforts through international round-robin testing, providing valuable and reliable data to support the development of dedicated safety standards for applications involving iron energy carriers.

Safety characteristics for gases, vapors, and dusts are typically found in databases, but contrary to gases and vapors, available data for dusts show a large spread and cannot be applied directly, as the results are less reproducible and difficult to implement in practice without conducting additional experiments. This is due to four influencing parameters that significantly affect their safety characteristics.

• Chemical properties (see 3.1)
• Moisture content (see 3.2)
• Particle size distribution (see 3.3)
• Particle morphology

In contrast, this study employs iron powders that were produced under identical conditions and characterized by the same chemical composition, moisture content, particle size distribution, and morphology. This minimizes variability and enables a clear assessment of whether the powders can be considered explosible under standardized testing, with safety aspects integrated from the beginning.

To date, driven by the increasing application of iron powders in emerging energy and industrial systems, research on the combustion behaviour of iron powders has expanded significantly. A substantial body of work has focused on single-particle combustion,²²^–²⁸ providing fundamental insights into ignition mechanisms, burning rates, and oxidation pathways under well-controlled conditions. Beyond single-particle combustion, further studies have addressed the behaviour of particles in clusters, suspensions and powder beds,²⁵^,²⁹^–³² as well as combustion modes, including laminar flame propagation.³³^–³⁸ These studies improve the understanding of collective combustion phenomena, such as particle–particle interactions and heat transfer within reacting clouds. However, they are typically conducted under controlled, idealized conditions, which limits their direct applicability to industrial dust-explosion scenarios. In contrast, studies directly addressing the explosion behaviour of iron dust and safety-related characteristics of iron dust, such as standard explosion parameters (e.g., maximum explosion pressure and rate of pressure rise), are reported,¹²^,³⁹^–⁴⁸ but remain comparatively limited. According to a previous study,³⁹ which summarises the available literature, iron dust is generally classified as St1, representing the lowest explosion class and meaning it is marginally explosible. It has been previously investigated and stated that the reproducibility of safety characteristic measurements of marginally explosible dusts, especially iron dusts, can be poor.⁴⁹^–⁵¹ Therefore, despite significant progress in understanding the fundamental mechanisms of combustion and flame configurations, a consistent, safety-oriented understanding of explosion behavior under standardised testing conditions — as required for industrial safety assessments — is still lacking.

After a first general approach to an a-priori-prediction of the safety characteristics of pure iron energy carriers,³⁹ this article represents a second step towards the development of a standardized iron powder as a dense energy carrier with the safety aspect implemented from the beginning. The aim is to establish an internationally harmonized understanding of the potential explosion hazards associated with this emerging technology and to define appropriate safety measures, fostering a holistic safety culture from the very beginning of its implementation. These data are also essential in a regulatory context, for example, within the framework of TDG/ADR for transport, REACH for substance registration and Safety Data Sheets, ATEX for the management of explosive atmospheres, and national regulations such as ICPE in France for industrial safety.

2.

Experimental

Experimental materials

2.1

The dust samples were provided by Pometon S.p.A. and Eindhoven University of Technology. Four types of iron dusts were tested: MT63, MT150, MT63reg, and MT150reg. The term “reg” indicates regenerated powders that have already undergone one combustion–reduction cycle.

The original MT63 and MT150 powders are produced by water atomization from molten scrap metal. For the regenerated powders, the reduction was performed in a mesh belt furnace operating at temperatures between 1000–1100°C with the addition of hydrogen, following combustion of the original MT63/150 powder in a 500-kWth pilot-scale industrial burner described in⁵² (Figures 2 and 3). After the reduction process, the iron is hammer-milled to produce a suitable particle size for combustion. Regenerated samples have a flaky shape due to the hammer-milling. The MT63reg sample has a d50 of about 50 μm, while the MT150reg sample would fall outside some standards because it is too coarse to be considered dust [1]. However, all powders have been successfully combusted in the pilot-scale MP500 burner and are therefore representative of iron powder for future power plants.

Figure 2. Picture of the reduction belt furnace.

Figure 3. Picture of the MP500 iron powder boiler.

All samples were characterized by their water content and particle size distribution (see Chapters 3.2 and 3.3). Although the surface morphology is not discussed further, the SEM images are provided in Figure 4 for reference. The tested iron dusts include:

• MT63: Iron dust freshly produced via water atomization, characterized by irregular, “potato-shaped” particles that are nonporous.
• MT150: Iron dust freshly produced via water atomization, characterized by irregular, “potato-shaped” particles that are nonporous, but coarser than MT63.
• MT63reg: Iron dust regenerated in a belt furnace following one combustion–reduction cycle, resulting in porous, thin, flake-like particles.
• MT150reg: Iron dust regenerated in a belt furnace following one combustion–reduction cycle, with porous, flake-like particles, but coarser than MT63reg.

Figure 4. Scanning Electron Microscopy (SEM) images showing a) MT63, b) MT150, c) MT63reg, and d) MT150reg.

Chemical composition

2.2

The chemical composition of the MT63 and MT150 powders was analyzed at the Bundesanstalt für Materialforschung und -prüfung (BAM) using energy-dispersive X-ray spectroscopy (EDX). The measurements were conducted in a Phenom XL G3 SEM (Thermo Fisher Scientific Inc.) with an acceleration voltage of 15 kV at a reduced pressure of 1–6 bar. For each sample, several measurements at different spots were carried out. In addition, line scans and mappings were recorded to investigate the elemental distribution. The level of impurities in the MT63reg and MT150reg powders can be assumed to be similar to the original MT63 and MT150, respectively, considering that no impurities were added during the combustion process. During the first combustion–reduction cycle, only minor changes in impurity concentrations can be expected to occur.⁶¹^,⁶² All powders were also tested for their crystalline composition using X-ray diffraction (XRD) and for their oxygen content using inert gas fusion (IGF) at Eindhoven University of Technology.

Experimental procedures

2.3

The dust samples were shipped to the Technical University of Leoben. There, the moisture contents, particle sizes, and safety characteristics were analyzed and subsequently packed and distributed to the other participating institutions:

• VSB – Technical University Ostrava, Czech Republic
• Slovak University of Technology in Bratislava, Slovakia
• ADINEX NV, Belgium
• Otto-von-Guericke University Magdeburg, Germany
• Bundesanstalt für Materialforschung und -prüfung (BAM), Germany
• Politecnico di Torino, Italy
• CMI – National Research Institute – Experimental Mine “Barbara”, Poland
• INERIS, Verneuil-en-Halatte, France

In general, round-robin particle characterization and explosion tests were carried out in nine facilities from eight different countries (see Figure 5).

Figure 5. Location of the nine participating facilities in central Europe.

As part of the coordinated effort, the institutes provided measurements of the moisture content, the particle size distribution, and the safety characteristics: explosibility, $p_{\max}$ , and $(dp / dt)_{\max}$ , or rather $K_{st}$ , for the four dusts, determined according to standardized procedures described in ISO 80079-20-2 and the EN 14034 series.⁵³^,⁵⁴^,⁵⁷

These safety characteristics were determined in standard 20-L spheres from different manufacturers. The main steps for the determination of safety characteristics are as follows (see also Figure 6):

1. Evacuation of the explosion vessel to a defined vacuum (0.4 bar absolute)
2. Pressurizing the dust chamber to a defined overpressure (21 bar absolute)
3. Opening of the fast-acting valve
4. Triggering the ignition source (2 × 1 kJ for explosibility, 2 × 5 kJ for $p_{\max}$ and $(dp / dt)_{\max}$ after a defined ignition delay time of 60 ms
5. Recording and evaluating the pressure development

Figure 6. Schematic of the 20 L-sphere. Taken from.⁶³

The dust is classified as non-explosible if no ignition test with an ignition energy of 2 × 1 kJ shows a pressure rise above 0.5 bar g over a wide dust concentration range⁵⁷ (see Figure 7).

Figure 7. Example of the measured pressure development for an iron dust sample: At a concentration of 750 g/m³ there is no ignition, whereas at 1500 g/m³ the pressure exceeds the threshold value of 0.5 bar g.

To determine $p_{\max}$ and $(dp / dt)_{\max}$ , the explosion overpressure $p_{ex}$ and the rate of pressure rise $(dp / dt)_{ex}$ are determined for each dust concentration and plotted as a function of dust concentration until a maximum value is obtained. The tests are performed for at least two consecutive concentrations on both sides of the maximum. This maximum corresponds to the maximum explosion overpressure $p_{ex}$ and the maximum rate of pressure rise $(dp / dt)_{\max}$ . If the procedure does not yield a clearly defined maximum, the test series is repeated in the region of the maximum. In such cases, the arithmetic mean of the peak values obtained in each test series is taken as maximum explosion overpressure $p_{ex}$ and $(dp / dt)_{\max}$ . The maximum explosion overpressure is processed with Equations 4 and 5 to determine $p_{\max} .$ ⁵⁴^,⁵⁸ The $K_{st}$ value is calculated from $(dp / dt)_{\max}$ according to Equation 6.⁵³

(4)

p_{max} = 5.5 ∙ \frac{p_{ex, max} - p_{ci}}{5.5 - p_{ci}}

(5)

p_{ci} = \frac{1.6 ∙ E_{i}}{10000}

(6)

K_{St} = (\frac{dp}{dt})_{max} * V_{TestVessel}^{1 / 3}

Although the test procedure is clearly defined in the standards, certain degrees of freedom in their execution, combined with the specific characteristics of iron dusts, can lead to deviations in the test results.

First, iron has a high density, and flame propagation can be observed at concentrations that are higher than for other dusts (e.g., coal or organic dusts such as flour). The dust concentrations in EN 14034–1/2 are not strictly predefined; instead, a recommendation is given: “The dust concentration is varied over a wide range from 250 g/m³ to 1500 g/m³ (typically 250 g/m³, 500 g/m³, 750 g/m³, 1000 g/m³ and 1500 g/m³) and the pressure increase is measured”.⁵³^,⁵⁴ It is further stated that in the case of high-density materials such as metals, higher concentrations (e.g., up to 2500 g/m³) are permitted. However, the testing procedure has practical limitations due to the need to inject the dust into the vessel using an overpressure (21 bar absolute) through an outlet valve and a rebound nozzle. The pressure increase in the sphere is typically around 0.6 bar, within a permissible range of 0.55–0.75 bar. Variations within this range may introduce small differences in the initial conditions.

At high dust concentrations and for high-density materials such as iron dust, not all of the injected air expands uniformly into the vessel before ignition. At concentrations above 1000 g/m³, the injection pressure is no longer sufficient to convey all dust into the sphere while still ensuring an initial pressure of 1 bar abs. As a result, ignition can occur at pressures below atmospheric conditions, leading to an underestimation of the measured pressure rise. Several approaches can be considered to address this limitation; however, most of them are restricted by standard test procedures:

• Increasing the pressure in the dust container to pressures above 21 bar (not allowed according to the standard)
• Increasing the initial pressure in the sphere before injection (also not allowed according to the standard)
• Testing the dust at pressures below ambient pressure (allowed but not recommended)
• Placing some of the dust sample or all of it in the container (allowed but not commonly applied; this also influences all safety characteristics⁶⁴^,⁶⁵)

In addition to these experimental constraints, the application of correction methods can further influence the interpretation of the results. Equation 4 stretches the measured values from a range of 5.5 − p_ci to 5.5 (p_ci is the pressure increase due to the igniters), which leads to an increase in the scattering of 41%. This means that small differences between individual measurements become amplified in the corrected values. This amplification should be considered when interpreting the results, since it may introduce additional variability that is not present in the raw measurements.

While the above aspects relate to the testing procedure, additional complexity arises from the intrinsic combustion behavior of iron dust. Iron dust is known for its relatively slow ignition behavior. It is often observed that a first pressure increase, caused by the igniters, is followed by a delayed second increase when iron dusts are tested.⁴¹^,⁴⁴ When the rate of pressure increase is determined automatically by the software, the first slope is used to determine the rate of pressure rise, while the iron itself reacts at a slower rate (see Figure 8).

Figure 8. Example of the measured pressure development for MT63 at a concentration of 2000 g/m³: difference in automatic and manual determination of $dp / dt$ .

Given the possible deviations caused by different determination techniques, all raw data were collected and analyzed in a consistent way. In this article, the rate of pressure rise is always determined from the second slope, not the first, which is allowed according to the American standard,⁵⁸ but not stated in the European standard.⁵³ For datasets where only values based on the first peak were initially reported, the corresponding values were re-evaluated by the authors using the second pressure rise peak from the available raw data.

3.

Results and discussion

Chemical composition

3.1

The results of the chemical analysis for the powders are shown in Tables 2 and 3.

Table 2. EDX analysis for MT63 and MT150, averaged values from 14, respectively 11 analyses.

The full analysis with all tests for each sample can be found in the Supplementary Material.

	Atomic fraction in %	Weight fraction in %
MT63
Iron	89.2	95.9
Oxygen	9.5	3.1
Aluminum	0.4	0.2
Silicon	0.3	0.1
Copper	0.3	0.4
Molybdenum	0.1	0.2
Sulfur	0.1	0.0
MT150
Iron	77.5	91.5
Oxygen	21.5	7.8
Aluminum	0.3	0.2
Silicon	0.5	0.3
Chromium	0.1	0.1

Table 3. Crystalline composition and oxygen mass fraction of the four powders.

No other compounds were found in the XRD pattern.

Powder	Fe [wt%]	FeO [wt%]	O wt% by XRD	O wt% by IGF
MT63	98.8	1.2	0.3	0.6
MT150	97.3	2.7	0.6	0.9
MT63reg	82.6	17.4	3.9	4.5
MT150reg	96.7	3.3	0.3	2.3

The EDX analysis data show that, as expected, the samples consist mainly of iron. However, MT63 and MT150 also contain significant amounts of oxygen, which indicates that some of the iron is present in oxidized form. The EDX mappings in the supplementary material, which show the distribution of elements, reveal that, unlike iron, oxygen is not distributed evenly throughout the sample, but rather occurs in distinct oxygen accumulations. This suggests the presence of discrete iron oxide accumulations within the powders.

Apart from oxygen, the samples contain no significant impurities. Although other elements such as copper and molybdenum were detected, their fractions are in very low range of less than 1 per cent.

The XRD measurements, conducted at TU Eindhoven, also show high values of iron and little to no impurities.

Moisture content

3.2

All facilities tested the dusts for their moisture content. The moisture content was analyzed using the DIN EN ISO 18134-1 method.⁶⁶ The results showed moisture contents for all samples ranging from 0.0 to 0.5 wt%, which is notably low and therefore neglected during further evaluation. The variations in the results presumably arise from differences in the testing methods, environmental conditions during transportation, or the relative humidity of the surrounding air in the laboratory during testing.

The values are so low that their impact on the safety characteristics can be considered negligible, as the mass of water is insufficient to act as a heat sink or influence the dispersion of the powder. Most investigations on the influence of moisture content on the explosion characteristics of dusts deal with values of several percent and higher, not below 1 wt%.⁶⁷^,⁶⁸

Particle size distribution

3.3

All facilities analyzed the particle size distribution using the laser diffraction method (ISO 13320:2020)⁶⁹ or mechanical sieving (ÖNORM EN 15415–1).⁷⁰ The characteristic particle size parameters $d_{10}$ , $d_{50}$ , and $d_{90}$ reported by the laboratories are summarized in Table 4 and Figure 9. Figure 9 presents the statistical distribution of particle sizes (D10, D50, D90) for all investigated powder samples. The graph was constructed using data obtained from both laser diffraction and mechanical sieving methods. The corresponding numerical values are summarized in Table 4, where the results are additionally distinguished based on the measurement technique (sieving and laser analysis).

Table 4. Characteristic particle size parameters (d10, d50, d90) reported by the participating laboratories.

Sample/Lab. number	MT63			MT150			MT63reg			MT150reg
D, μm	d₁₀	d₅₀	d₉₀	d₁₀	d₅₀	d₉₀	d₁₀	d₅₀	d₉₀	d₁₀	d₅₀	d₉₀
1	6	48	153	20	58	139	1	31	69	0	66	254
2	27	54	96	41	87	154	-	-	-	-	-	-
5	24	51	90	40	85	141	13	38	69	33	80	139
6	26	47	84	34	72	133	12	39	67	29	69	137
7	26	52	97	32	74	145	20	42	73	37	83	168
8	26	51	86	31	68	128	-	-	-	33	73	151
9	25	52	126	32	76	169	14	40	69	29	72	157
Min. value	6	47	84	20	58	128	1	31	67	0	66	137
Max. value	27	54	153	41	87	169	20	42	73	37	83	254
Median	26	51	96	33	74	141	13	39	69	31	72	154
3*	15	36	70	20	57	108	15	32	54	25	60	113
4*	17	40	68	24	63	110	19	36	58	28	63	116

Figure 9. Statistical distribution of particle sizes (D10, D50, D90) for all investigated powder samples (MT63, MT150, MT63reg, MT150reg).

The boxplots represent the interquartile range (25th–75th percentile), with the median indicated by a horizontal line inside the box. Whiskers denote the minimum and maximum values. The marker (•) indicates the mean particle size.

The results demonstrated a similar size distribution with only minor differences between laboratories for MT63, MT150, and MT63reg (see Table 4). Larger deviations are observed for MT150reg, particularly for $d_{90}$ , which indicates the presence of coarse particles in some subsamples and suggests a lower homogeneity of this powder.

In cases where both measurement methods were applied, the values obtained by laser diffraction tend to be higher than those obtained by sieve analysis, particularly for the upper percentile $d_{90}$ . This difference reflects the fundamental differences between the methods. Laser diffraction determines an equivalent spherical volume diameter and is sensitive to irregularly shaped particles or agglomerates, which can increase the upper percentiles of the distribution. In contrast, sieve analysis is based on the smallest particle dimension that allows passage through the mesh opening and may therefore underestimate the contribution of irregular or weakly agglomerated coarse particles.⁷¹

For the interpretation of the explosibility data, it is essential to have reliable information on the particle size range. At the same time, it must be emphasized that different characterization techniques come with inherent limitations. Thus, the observed deviations between laboratories do not indicate different samples but rather reflect method-specific sensitivities (e.g., laser diffraction vs. sieve analysis).

Safety characteristics

3.4

3.4.1 Explosibility

First, all samples were tested for explosibility according to ISO 80079-20-2 using two 1 kJ ignitors. If an explosion was detected, further tests were conducted to determine explosion parameters according to EN 14034–1/2,⁵³^,⁵⁴ using two 5 kJ ignitors.

The explosibility results for all four dusts are summarized in Table 5. The results of the round-robin test indicate variability in the classification of dust explosibility among laboratories. MT63 was considered non-explosible by six out of nine laboratories, and MT150 by seven out of nine. MT63reg was classified as non-explosible by only one laboratory, whereas MT150reg was classified as non-explosible by six out of nine laboratories. A comparison of the particle size distributions reported by each laboratory (Table 4) with the corresponding explosibility results (Table 5) did not reveal any clear correlation between the characteristic particle size parameters (d₁₀, d₅₀, d₉₀) and the observed explosion behavior.

Table 5. Explosibility of all four dust samples.

N.	MT63	MT150	MT63reg	MT150reg
1	Explosible	Non-explosible	Explosible	Non-explosible
2	Explosible	Explosible	Explosible	Non-explosible
3	Non-explosible	Non-explosible	Explosible	Non-explosible
4	Non-explosible	Non-explosible	Explosible	Non-explosible
5	Non-explosible	Non-explosible	Explosible	Explosible
6	Non-explosible	Non-explosible	Explosible	Explosible
7	Explosible	Explosible	Explosible	Explosible
8	Non-explosible	Non-explosible	Explosible	Non-explosible
9	Non-explosible	Non-explosible	Non-explosible	Non-explosible

The observed variations in explosibility results between laboratories may be attributed to the degrees of freedom in the execution of the tests allowed by the standardized procedures, as described in detail in the methodology section. On the one hand, different concentration ranges were considered for the determinations (e.g., up to 1500 or 2500 g/m³). On the other hand, the initial pressures at which ignition occurred were not always consistent. To overcome the challenge of conveying high-density dust at elevated concentrations (above 1000 g/m³) into the sphere, some laboratories accepted ignition at reduced initial pressures, which may lead to false-negative results. The detailed results of all dust explosion tests conducted by all participating laboratories are provided in the Supplemental Material.

It can be observed that the regenerated samples (MT63reg, MT150reg) showed positive explosion results more frequently than the non-regenerated samples (MT63, MT150). Since only minor differences in particle size distribution were observed, this behavior is more likely associated with differences in particle morphology, as the more porous and flake-like structure increases the specific surface area and thus enhances reactivity.⁷²

3.4.2 Explosion severity

For the cases where an explosion was recorded, the safety characteristics such as the maximum explosion pressure $p_{\max}$ , the maximum rate of pressure rise $(dp / dt)_{\max}$ , and the volume-normalized deflagration index $K_{st}$ were determined (Tables 6–9).

Table 6. Evaluation of MT63 explosion indices based on data reported by three participants.

Pressure values were rounded to one decimal place, whereas rate of pressure rise values and percentages were rounded to integers.

Parameter [N. participant]	Mean value ± standard deviation	Relative standard deviation [%]
$p_{ex}$ in bar g	3.3 ± 0.6	±19
$p_{ex}$ in bar g (Lab 1)	3.8 ± 0.6	±17
$p_{ex}$ in bar g (Lab 2)	3.6 ± 0.1	±3
$p_{ex}$ in bar g (Lab 7)	2.6 ± 0.1	±6
$p_{\max}$ in bar g	2.5 ± 0.9	±38
$p_{\max}$ in bar g (Lab 1)	3.1 ± 0.9	±29
$p_{\max}$ in bar g (Lab 2)	2.9 ± 0.2	±5
$p_{\max}$ in bar g (Lab 7)	1.4 ± 0.2	±14
$(dp / dt)_{\max}$ in bar/s	26 ± 7	±27
$(dp / dt)_{\max}$ in bar/s (Lab 1)	26 ± 8	±31
$(dp / dt)_{\max}$ in bar/s (Lab 2)	19 ± 5	±27
$(dp / dt)_{\max}$ in bar/s (Lab 7)	33 ± 6	±17
$K_{st}$ in bar·m/s	7 ± 2	±29
$K_{st}$ in bar·m/s (Lab 1)	7 ± 2	±31
$K_{st}$ in bar·m/s (Lab 2)	5 ± 1	±27
$K_{st}$ in bar·m/s (Lab 7)	9 ± 2	±17

Uncertainty in the results also appeared during the determination of the explosion characteristics. According to the standards, duplicate measurements within one laboratory should agree within 5% for the maximum explosion pressure $p_{\max}$ , while the reproducibility between laboratories should be within 10%.⁵⁴ For the deflagration index $K_{st}$ or the maximum rate of pressure rise $(dp / dt)_{\max}$ , standards specify a repeatability and reproducibility of 30%.⁵³ However, as shown in Tables 6–9, the observed deviations partially exceed these limits. It should be emphasized that the particle size distributions of the tested samples were very similar (Table 4) between the laboratories, indicating that the observed differences are unlikely to originate from variations in particle size. A plausible explanation for this behavior lies in the specific combustion characteristics of iron. Iron is often classified as a St 1 Class dust, indicating weak explosibility and typically resulting in poorer reproducibility,⁴⁹^–⁵¹ as discussed above.

The standard deviation for the maximum explosion pressure ( $p_{\max}$ ) between the laboratories is a lot higher than 10% (Tables 6–9). This can be attributed to the data evaluation procedure specified in both American and European standards, where the correction applied according to Equations (4) and (5) increases the scattering by approximately 41%, potentially leading to higher deviations than actually measured. In addition, the signal-to-noise ratio is inherently low: a typical scattering of 0.5 bar corresponds to a much larger relative deviation when the maximum pressure is 2 bar for weakly reactive iron dust than when it is 10 bar for other commonly tested dusts.

Moreover, it should be noted that the actual pressure increase caused solely by the ignitors ( $p_{ci}$ ) is not always exactly 1.6 bar but generally assumed (Equation 5). This phenomenon was described in.⁷³ Laboratories often rely on manual assessment of their ignitors, and the resulting values can deviate; for example, instead of 1.6 bar, values as low as 1.1 bar are reported.⁷³ The scattering of the pressure increase caused solely by the ignitors from one batch to another is about 0.1 bar and has a relatively stronger effect in case of weak explosiveness.⁷⁴ This re-evaluation contributes to the discrepancies observed between laboratories in terms of $p_{\max}$ .

For this reason, the row data were additionally processed by the authors and the corresponding $p_{ex}$ values were included in the Tables 6–9. As can be seen, the standard deviation between laboratories decreases after this re-evaluation. For MT63 and MT150, the standard deviation between laboratories exceeds the acceptable limits, but remains lower than that observed for $p_{\max}$ . For MT63reg and MT150reg, the deviations between laboratories remain within the acceptable range.

The deviations observed between laboratories can be attributed to the remaining degrees of freedom in the execution of the tests, as described in the methodology section, as well as to specific characteristics of iron dust. In particular, the selection of dust concentrations for testing was not uniform across laboratories. Since the concentration range is not strictly predefined and is adapted based on the occurrence of the first positive explosibility result, different concentration ranges were applied, which influenced the subsequent determination of explosion parameters. In addition, practical limitations related to the injection of high-density dust at elevated concentrations further contributes to these variations.

Tables 6–9 also provide the standard deviation with in individual laboratories. It should be noted that these values were calculated only for laboratories that performed multiple test series. In these cases, the standard deviation was determined based on the maximum values obtained from the individual test series within the same laboratory.

As shown in Tables 6–9, the standard deviation within individual laboratories exceeds the 5% limit in several cases. This indicates that significant variability is present not only between laboratories but also within individual laboratories. Such deviations highlight the influence of the dust injection system and its limited repeatability on the measured results. This effect becomes particularly significant at low pressure levels, where even small absolute deviations (e.g., around 0.1 bar) correspond to relative differences exceeding 5%.

With regard to the differences observed in $(dp / dt)_{\max}$ , all samples exhibited values that did not exceed 50 bar/s. As shown in Tables 6–9, the standard deviation between laboratories for the rate of pressure rise remains within the acceptable limits for MT63 and MT63reg. For MT150 and MT150reg, however, the deviations reach approximately 39%, which is around 9% above the recommended limit, but not considered critical. At the same time, the deviations within individual laboratories remained within the 30% range in all cases. This is particularly noteworthy, as it indicates a good repeatability of the measurements for the second pressure rise within individual laboratories.

A separate comparison of the deflagration index $K_{st}$ was performed; however, as this parameter is directly derived from the rate of pressure rise by applying a scaling factor, a separate discussion does not provide additional insight. The corresponding values are included in Tables 6–9.

Figures 10–19 show the individual experimental results $p_{ex}$ and $(dp / dt)_{ex}$ plotted as data points. For improved readability, lines representing the average trend are additionally included, based on the average values at each concentration or on single test results, depending on the data reported by each laboratory.

Figure 10. Explosion overpressure $p_{ex}$ for MT63 as a function of dust concentration measured by Lab. 1, 2, 7.

Based on theoretical considerations, these plots are generally expected to exhibit a curve with a local maximum at a certain dust concentration, after which maximum values should decrease due to the excess thermal mass of unreacted powder.⁴¹^,⁷⁵ However, when examining the $p_{ex} / (dp / dt)$ –concentration plots obtained in this work, a different behavior can be observed. Instead of a pronounced maximum followed by a decrease, the curves rather show an increase in pressure values followed by a plateau with some fluctuations.

An important aspect in this context to consider is that the maximum explosion pressure for iron-based dust is reported to occur at relatively high dust concentrations (around 2500 g/m³),⁴⁴ whereas the oxygen available in the 20-L vessel can theoretically oxidize only approximately 660 g/m³ (Fe₂O₃), 740 g/m³ (Fe₃O₄), or 990 g/m³ (FeO).

This discrepancy arises because the theoretical stoichiometric concentration assumes that the entire amount of added dust is uniformly dispersed within the 20-L spherical vessel. In reality, this condition is difficult to achieve, especially for dense dusts at high concentrations. This behavior suggests that only a fraction of the nominal dust mass participates in heat absorption. This can be attributed to several factors, including incomplete material injection into the vessel, as well as strongly non-uniform particle dispersion within the sphere. All these aspects contribute to the fact that iron dust exhibits significant deviations in the measured explosibility parameters.

Figures 10 and 11 present the explosion overpressure $p_{ex}$ and the maximum rate of pressure rise $(dp / dt)_{ex}$ for MT63 obtained in three laboratories. It should be noted that the selection of dust concentrations for further testing depends on the concentration at which a positive explosion test is obtained using 2 kJ igniters. For example, a positive result was observed at 2000 g/m³ in the first laboratory. Consequently, the first laboratory focused on concentrations above 2000 g/m³, while the seventh laboratory limited its testing to concentrations up to approximately 1500 g/m³. In contrast, the second laboratory investigated a broader range, covering concentrations from 750 to 3000 g/m³. As discussed in the methodology section, testing at such high concentrations is limited by the 20-L apparatus’s injection capacity. Therefore, the differences in the selected concentration ranges for the four powders are due to these practical constraints.

Figure 11. Rate of pressure rise $(dp / dt)_{ex}$ for MT63 as a function of dust concentration measured by Lab. 1, 2, 7.

A comparison of the MT63 results shows that the first and the second laboratories exhibit relatively similar behavior in $p_{ex}$ . In contrast, the seventh laboratory reports lower maximum explosion pressure, which leads to higher standard deviation between laboratories (Table 6). With regard to the rate of pressure rise, the opposite trend is observed, as the highest values were observed in the seventh laboratory. Standard deviation for highest values between laboratories in $(dp / dt)_{ex}$ is in acceptable range, as well as inside each laboratory.

Figures 12 and 13 present the results obtained for MT150. The general trend is similar to that observed for MT63. As described previously, different dust concentrations were tested. The second laboratory investigated a wider concentration range (1000–4000 g/m³) with larger step intervals, whereas the seventh laboratory focused on concentrations up to 1500 g/m³ with smaller step sizes. The measured explosion overpressures are generally higher in the 2^nd laboratory, whereas the 7^th laboratory reports higher values of the maximum rate of pressure rise. The repeatability of the maximum pressure within the 2^nd laboratory remains within the acceptable range, whereas within the 7^th laboratory exceeds the limit. Between-laboratory reproducibility is outside the acceptable range. The repeatability of the maximum rate of pressure rise within individual laboratories remains within the acceptable range according to the standard. However, when comparing the maximum values between laboratories, the standard deviation reaches approximately 39%, slightly exceeding the recommended limit of 30% (Table 7).

Figure 12. Explosion overpressure $p_{ex}$ for MT150 as a function of dust concentration measured by Lab. 2, 7.

Figure 13. Rate of pressure rise $(dp / dt)_{ex}$ for MT150 as a function of dust concentration measured by Lab. 2, 7.

Table 7. Evaluation of MT150 explosion indices based on data reported by two participants.

Pressure values were rounded to one decimal place, whereas rate of pressure rise values and percentages were rounded to integers.

Parameter [N. participant]	Mean value ± standard deviation	Relative standard deviation [%]
$p_{ex}$ in bar g	2.9 ± 0.8	±27
$p_{ex}$ in bar g (Lab 2)	3.4 ± 0.1	±3
$p_{ex}$ in bar g (Lab 7)	2.3 ± 0.4	±16
$p_{\max}$ in bar g	1.9 ± 1.1	±57
$p_{\max}$ in bar g (Lab 2)	2.6 ± 0.2	±7
$p_{\max}$ in bar g (Lab 7)	1.1 ± 0.6	±52
$(dp / dt)_{\max}$ in bar/s	24 ± 9	±39
$(dp / dt)_{\max}$ in bar/s (Lab 2)	17 ± 1	±9
$(dp / dt)_{\max}$ in bar/s (Lab 7)	30 ± 8	±27
$K_{st}$ in bar·m/s	7 ± 2	±33
$K_{st}$ in bar·m/s (Lab 2)	5 ± 0.4	±9
$K_{st}$ in bar·m/s (Lab 7)	8 ± 3	±33

Figures 14 and 15 present the results obtained for MT150reg. In this case, both laboratories tested within largely comparable concentration ranges. It can be observed that at concentrations of 1000 and 1250 g/m³, the values for both explosion parameters differ significantly between the laboratories. Nevertheless, when considering the average of the maximum values for each parameter in each series, the between-laboratory reproducibility remains within acceptable limits for the maximum explosion pressure, while it is slightly above the recommended range for the maximum rate of pressure rise (Table 8).

Figure 14. Explosion overpressure $p_{ex}$ for MT150reg as a function of dust concentration measured by Lab. 5, 7.

Figure 15. Rate of pressure rise $(dp / dt)_{ex}$ for MT150reg as a function of dust concentration measured by Lab. 5, 7.

Table 8. Evaluation of MT150reg explosion indices based on data reported by two participants.

Pressure values were rounded to one decimal place, whereas rate of pressure rise values and percentages were rounded to integers.

Parameter [N. participant]	Mean value ± standard deviation	Relative standard deviation [%]
$p_{ex}$ in bar g	3.0 ± 0.1	±2
$p_{ex}$ in bar g (Lab 5)	3.0 ± 0.1	±2
$p_{ex}$ in bar g (Lab 7)	2.9 ± 0.4	±13
$p_{\max}$ in bar g	1.9 ± 0.1	±7
$p_{\max}$ in bar g (Lab 5)	2.0 ± 0.1	±3
$p_{\max}$ in bar g (Lab 7)	1.8 ± 0.5	±28
$(dp / dt)_{\max}$ in bar/s	28 ± 11	±39
$(dp / dt)_{\max}$ in bar/s (Lab 5)	20 ± 0	±0
$(dp / dt)_{\max}$ in bar/s (Lab 7)	35 ± 4	±12
$K_{st}$ in bar·m/s	7 ± 3	±40
$K_{st}$ in bar·m/s (Lab 5)	5 ± 0	±0
$K_{st}$ in bar·m/s (Lab 7)	9 ± 1	±12

Figures 16–19 show the explosion overpressure ( $p_{ex}$ ) and the rate of pressure rise for MT63reg. As summarized in Table 9, all deviations remain within the acceptable range. Nevertheless, it can be observed that the pressure peaks occur at different dust concentrations. As this sample was classified as explosible by the majority of the participating laboratories, a larger dataset is available compared to the other powders. Over a wide concentration range (500–5000 g/m³), a consistent overall trend can be observed, with no clear decline in the measured values at higher concentrations. Instead, both parameters increase at lower concentrations and then reach a plateau with some fluctuations. This behavior highlights the limitations of the 20-L sphere when testing marginally reactive dusts, where further increases in dust concentration do not result in lower explosion severity.

Figure 16. Explosion overpressure p_ex for MT63reg as a function of dust concentration measured by Lab. 1, 4, 5, 7.

Figure 17. Explosion overpressure p_ex for MT63reg as a function of dust concentration measured by Lab. 2, 3, 6.

Figure 18. Rate of pressure rise $(dp / dt)_{ex}$ for MT63reg as a function of dust concentration measured by Lab. 2, 3, 4, 6.

Table 9. Evaluation of MT63reg explosion indices based on data reported by seven participants.

Pressure values were rounded to one decimal place, whereas rate of pressure rise values and percentages were rounded to integers.

Parameter [N. participant]	Mean value ± standard deviation	Relative standard deviation [%]
$p_{ex}$ in bar g	3 ± 0.2	±7
$p_{ex}$ in bar g (Lab 1)	3.2 ± 0.1	±2
$p_{ex}$ in bar g (Lab 2)	3.0 ± 0.03	±0.9
$p_{ex}$ in bar g (Lab 3)	3.2	-
$p_{ex}$ in bar g (Lab 4)	3.1	-
$p_{ex}$ in bar g (Lab 5)	2.9 ± 0.1	±2
$p_{ex}$ in bar g (Lab 6)	3.0	-
$p_{ex}$ in bar g (Lab 7)	2.6 ± 0.1	±3
$p_{\max}$ in bar g	2.1 ± 0.4	±18
$p_{\max}$ in bar g (Lab 1)	2.3 ± 0.1	±3
$p_{\max}$ in bar g (Lab 2)	2.0 ± 0.1	±3
$p_{\max}$ in bar g (Lab 3)	2.6	-
$p_{\max}$ in bar g (Lab 4)	2.1	-
$p_{\max}$ in bar g (Lab 5)	1.8 ± 0.2	±8
$p_{\max}$ in bar g (Lab 6)	2.4	-
$p_{\max}$ in bar g (Lab 7)	1.5 ± 0.1	±4
(dp/dt)_max in bar/s	37 ± 10	±27
(dp/dt)_max in bar/s (Lab 1)	50 ± 0	±0
(dp/dt)_max in bar/s (Lab 2)	30 ± 0	±0
(dp/dt)_max in bar/s (Lab 3)	53	-
(dp/dt)_max in bar/s (Lab 4)	35	-
(dp/dt)_max in bar/s (Lab 5)	27 ± 3	±11
(dp/dt)_max in bar/s (Lab 6)	32	-
(dp/dt)_max in bar/s (Lab 7)	35 ± 0	±0
$K_{st}$ in bar·m/s	10 ± 3	± 28
$K_{st}$ in bar·m/s (Lab 1)	14 ± 0	± 0
$K_{st}$ in bar·m/s (Lab 2)	8 ± 0	± 0
$K_{st}$ in bar·m/s (Lab 3)	14	-
$K_{st}$ in bar·m/s (Lab 4)	9	-
$K_{st}$ in bar·m/s (Lab 5)	7 ± 1	± 11
$K_{st}$ in bar·m/s (Lab 6)	9	-
$K_{st}$ in bar·m/s (Lab 7)	9 ± 0	± 0

Figure 19. Rate of pressure rise $(dp / dt)_{ex}$ for MT63reg as a function of dust concentration measured by Lab. 1, 5, 7.

Figure 20. SIE-Chart: Explosion pressures for several iron dusts against the stated d50, Data taken from.³⁹^,⁴⁰^,⁷⁶

Overall, all powders exhibited similar levels of explosion overpressure ( $p_{ex}$ ), with average maximum values of approximately 3 bar, regardless of whether the material was regenerated or not. Although it was expected that the regenerated samples would show higher explosion severity due to changes in particle morphology and structure,⁷² this effect is not clearly observed in the results.

The same applies to the rate of pressure rise $(dp / dt)_{ex}$ , for which the average maximum values are around 30 bar/s across all samples, with no significant differences between regenerated and non-regenerated powders. This suggests that, despite morphological changes, the overall explosion severity remains comparable.

In summary, the round-robin test revealed variability in the determination of explosion parameters for weakly explosive iron dust, despite consistent sample properties and standardized procedures. These findings highlight the limitations of the current testing procedure, particularly for low-reactivity dusts. They show that variations in the safety characteristics of different iron dust samples with different properties may be overshadowed by the limited reproducibility of the test methods.

As long as these limitations in the explosion standard persist, the establishment of general safety concepts for a standardized, non-explosive iron energy carrier — intended for large-scale distribution across Europe — remains a challenge. One way to overcome this difficulty is to classify all iron dusts as St 1 ( $K_{st}$ < 200) or to use an a-priori approach as suggested previously³⁹ (see Figure 20) in which the expected value is the linear regression for all the data points, the upper limit is the linear plot between the highest observed data points. The theoretical value was calculated with the assumption that a full conversion from Fe to Fe₂O₃ at the stoichiometry with air takes place when the particles are infinitely small and are evenly suspended in air. Further assumptions/values for the calculation were a constant heat capacity for iron of 0,55 kJ/kg/K and 1,01 kJ/kg/K for air, and a latent heat of melting for iron of 13,8 kJ/kg.

Further work is required to populate this plot with additional data, but eventually it may allow a classification without the need for further testing of iron energy carriers. Simultaneously, a better understanding of dispersion, turbulence, and combustion/explosion in a 20-liter vessel - for instance, through comparison with numerical simulations, as demonstrated in³²- could help explain the observed variation and support the development of an improved standard.

4.

Conclusion

Four powders currently used in pilot-scale reactors for the iron powder energy carrier concept were tested for their explosion characteristics by performing a round-robin experiment. The round-robin test confirmed that all nine participating laboratories received and analyzed dust samples with consistent physical properties, particularly in terms of moisture content and particle size distribution. Standardized testing procedures were applied for the initial screening for explosibility and the subsequent determination of the explosion parameters, according to ISO 80079-20-2⁵⁷ and EN 14034–1/2.⁵³^,⁵⁴

The results revealed variability in the classification of dust explosibility, where disagreement between laboratories was evident even though all laboratories followed the same standard procedures. The reasons for these variations may arise from differences in the dust concentrations tested or lower starting pressures. A scenario in which this may cause severe problems is easy to imagine: a producer tests the dust in one laboratory and it is classified as non-explosible, then ships it to another company for application. If an accident occurs, the second company may test the dust again in another laboratory, which then classifies it as explosible—despite both following the ISO 80079-20-2⁵⁷ standard. In addition to differing classifications regarding explosibility, deviations were observed in the values of the maximum explosion pressure and the maximum rate of pressure rise, partially exceeding the tolerances defined in the standards.

These discrepancies can be attributed to the specific combustion behavior of weakly reactive iron dusts. For the maximum explosion pressure, the main contributing factors are:

• A lower signal-to-noise ratio, with a typical scattering of 0.5 bar on a maximum pressure of approximately 2 bar for weakly reactive iron dusts, compared to about 10 bar for other commonly tested dusts.
• Correction of the measured pressures to the standardized reference pressure below 5.5 bar g. This correction increases the scattering by 41%, which may lead to higher deviations than actually measured.
• The inherent scattering of the overpressure generated by the ignitors, typically around 0.1 bar from one igniter batch to another.
• Differences in the initial pressure arising from friction in the injection process.⁵⁶
• The concentration limits unto which the samples were tested.

For the maximum rate of pressure rise, the dual pressure-rise pattern should be addressed in the standards, and it should be clearly indicated which pressure rise has to be taken. Inconsistencies in the evaluation methods (manual vs. automated) otherwise lead to a wide spread of the data.

The findings reveal limitations in current explosion testing procedures for low-reactivity dusts, highlighting the need for methodological improvements and clearer evaluation criteria. However, the round-robin test confirmed that all four powders investigated were classified either as non-explosible or as Class 1 marginally explosible dust. This provides a clear picture of the level of explosion protection measures that need to be considered for the safe use of iron powders in energy-carrier applications. Despite the discrepancies in the explosion parameters measured between laboratories, the overall results demonstrate that the explosibility of iron powders remains low.

This round-robin test, which represents one of the very few inter-laboratory studies on iron powders, contributes both to improving our understanding of iron dust explosibility and to supporting future refinement of safety standards.

An a priori approach, such as the one presented in this study, could provide a viable solution for the characterization of pure dusts, reducing uncertainties and complementing standardized methods. Together, the results not only show where methodologies can be improved but also provide confidence that iron powders can be handled safely as a recyclable energy carrier in large-scale applications, provided that appropriate protective measures are implemented.

Ethics and consent

Ethical approval and consent were not required.

Data availability statement

Underlying data

Zenodo: CESAR. IRON ROUND ROBIN Semenova and Spitzer 2026. CESAR. https://doi.org/10.5281/zenodo.19726825.⁷⁷

The project contains the following underlying data:

• EDX_MT150_250429_neu_2.pdf contains the EDX measurements for MT150.
• EDX_MT63_250429_neu_2.pdf contains the EDX measurements for MT63.
• Expl.tests.pdf contains the test results from all eight facilities.
• RoundRobin_Explosion_Parameters.xlsx

Data are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).

Acknowledgments

The corresponding author would like to thank Paul Geoerg for drawing the map of Europe.

Courtesy to Tom Siebelt and Bart van Overbeeke for providing the photograph displayed in Figure 3.

References

1. Bergthorson JM, Goroshin S, Soo MJ, et al.: Direct combustion of recyclable metal fuels for zero-carbon heat and power. Applied Energy. 2015; 160: 368–382. Publisher Full Text
2. Metalot: It’s time for Circular Energy: It’s time for Iron Power—Iron Power Program: Strategic Rationale.2025.
3. Neumann J, Da Rocha RC, Debiagi, et al.: Techno-economic assessment of long-distance supply chains of energy carriers: Comparing hydrogen and iron for carbon-free electricity generation. Appl Energy Combust Sci. 2023; 14: 100128. Publisher Full Text
4. Church C, Crawford A: Minerals and the Metals for the Energy Transition: Exploring the Conflict Implications for Mineral-Rich :Fragile States; 2020.
5. Bock S, Pauritsch M, Lux S, et al.: Natural iron ores for large-scale thermochemical hydrogen and energy storage. Energy Conversion and Management. 2022; 267: 115834. Publisher Full Text
6. Voitic G, Hacker V: Recent advancements in chemical looping water splitting for the production of hydrogen. RSC Advances. 2016; 6(100): 98267–98296. Publisher Full Text
7. Quitzow R, Zabanova Y: The Geopolitics of Hydrogen: Volume 1: European Strategies in Global Perspective. Cham: Springer Nature Switzerland; Imprint Springer; 2024.
8. Hafner M, Tagliapietra S: The geopolitics of the global energy transition. Cham: Springer Open; 2020.
9. Hydrogen Research Europe: Advancing Hydrogen Technologies – A roadmap for pioneering research projects. Hydrogen Research Europe. November 2025.
10. Geurts B, Beijer A: World’s first in fossil-free industrial heat – RIFT and Kingspan Unidek sign historic first commercial contract for Iron Fuel Technology. 2025.https://www.kingspangroup.com/en/news-insights/rift-and-kingspan-unidek-sign-historic-first-commercial-contract/
11. Boretti A, Castelletto S: Metal fuels for renewable energy storage and circular economy. Next Energy. 2025; 8: 100371. Publisher Full Text
12. Eckhoff RK: Dust explosions in the process industries: Identification, assessment and control of dust hazards. Gulf Professional Publishing; 2003.
13. Manilevich FD, Pirskyy Y, Kutsyi AV, et al.: Aluminium as an efficient energy storage substance for the catalysed generation of hydrogen from water. J Energy Storage. 2024; 96: 112748. Publisher Full Text
14. Kim W, Saeki R, Ueno Y, et al.: Effect of particle size on the minimum ignition energy of aluminum powders. Powder Technology. 2023; 415: 118190. Publisher Full Text
15. World Nuclear Association: Heat Values of Various Fuels. 2026.https://world-nuclear.org/
16. Kecojevic V, Vukotic I, Komljenovic D: Production, consumption and cost of energy for surface mining of bituminous coal. Min Eng. 2014; 66: 51–57.
17. Norman F, Berghmans J, Verplaetsen F: The Minimum Ignition Energy of Coal Dust in an Oxygen Enriched Atmosphere. Chem Eng Trans. 2013; 31.
18. Pedišius N, Praspaliauskas M, Pedišius J, et al.: Analysis of Wood Chip Characteristics for Energy Production in Lithuania. Energies. 2021; 14(13): 3931. Publisher Full Text
19. CROHMIQ: Minimum Ignition Energy Data by Product. 2026.https://crohmiq.com/minimum-ignition-energy-data-by-product/
20. DECHEMA: CHEMSAFE – Explosion Characteristics of Substances and Mixtures. 2026.https://dechema.de/en/chemsafe.html
21. Pan C, Wang X, Li G, et al.: Influences of a square and circular vent on gasoline vapor explosions and its propagation: Comparative experimental study. Case Studies in Thermal Engineering. 2021; 27: 101225. Publisher Full Text
22. Ogle RA: Dust Explosion Dynamics. Butterworth-Heinemann, Elsevier; 2016.
23. Bidabadi M, Mafi M: Time variation of combustion temperature and burning time of a single iron particle. International Journal of Thermal Sciences. 2013; 65: 136–147. Publisher Full Text
24. Panahi A, Chang D, Schiemann M, et al.: Combustion behavior of single iron particles-part I: An experimental study in a drop-tube furnace under high heating rates and high temperatures. Appl Energy Combust Sci. 2023; 13: 100097. Publisher Full Text
25. Mi X, Fujinawa A, Bergthorson JM: A quantitative analysis of the ignition characteristics of fine iron particles. Combustion and Flame. 2022; 240: 112011. Publisher Full Text
26. Fujinawa A, Thijs LC, Jean-Philyppe J, et al.: Combustion behavior of single iron particles, Part II: A theoretical analysis based on a zero-dimensional model. Applications in Energy and Combustion Science. 2023; 14: 100145. Publisher Full Text
27. Ning D, Shoshin Y, van Oijen JA , et al.: Burn time and combustion regime of laser-ignited single iron particle. Combustion and Flame. 2021; 230: 111424. Publisher Full Text
28. Ning D, Hazenberg T, Shoshin Y, et al.: Experimental and theoretical study of single iron particle combustion under low-oxygen dilution conditions. Fuel. 2024; 357: 129718. Publisher Full Text
29. Leshchevich VV, Penyazkov OG, Fedorov AV, et al.: Conditions and delay time of ignition of iron microparticles in oxygen. Journal of Engineering Physics and Thermophysics. 2012; 85: 148–154. Publisher Full Text
30. Vance FH, Scholtissek A, Nicolai H, et al.: Flame propagation modes for iron particle clusters in air — Part I: Transition from continuous to discrete propagation mode under weak convection effects. Combustion and Flame. 2024; 260: 113265. Publisher Full Text
31. Vance FH, Scholtissek A, Nicolai H, et al.: Flame propagation modes for iron particle clusters in air, Part II: Transition from continuous to discrete propagation mode under strong convection effects. Combustion and Flame. 2024; 265: 113199. Publisher Full Text
32. Fujinawa A, Mi X: Numerical studies on the propagation of iron dust flames in confinement. Proceedings of the Combustion Institute. 2024; 40(1–4): 105754. Publisher Full Text
33. Goroshin S, Tang F-D, Higgins AJ, et al.: Laminar dust flames in a reduced-gravity environment. Acta Astronautica. 2011; 68(7–8): 656–666. Publisher Full Text
34. Tang F-D, Goroshin S, Higgins A, et al.: Flame propagation and quenching in iron dust clouds. Proceedings of the Combustion Institute. 2009; 32(2): 1905–1912. Publisher Full Text
35. Tang F-D, Goroshin S, Higgins AJ: Modes of particle combustion in iron dust flames. Proceedings of the Combustion Institute. 2011; 33(2): 1975–1982. Publisher Full Text
36. Hazenberg T, Braig D, Fedoryk MA, et al.: Analyzing iron dust bunsen flames using numerical simulations. Fuel. 2026; 403: 136094. Publisher Full Text
37. Fedoryk M, Stelzner B, Harth S, et al.: Experimental investigation of the laminar burning velocity of iron-air flames in a tube burner. Applications in Energy and Combustion Science. 2023; 13: 100111. Publisher Full Text
38. van Gool C , Hazenberg T, van Oijen JA , et al.: Numerical determination of iron dust laminar flame speeds with the counter-flow twin-flame technique. Combust Flame. 2024; 266: 113524. Publisher Full Text
39. Spitzer sen SH, Stevens NC, Krietsch A, et al.: A first step towards a standardized iron fuel: Safety characteristics of several pure iron powders. Fuel. 2025; 400: 135824. Publisher Full Text
40. Danzi E, Marmo L: Dust explosion risk in metal workings. Journal of Loss Prevention in the Process Industries. 2019; 61: 195–205. Publisher Full Text
41. Danzi E, Pio G, Marmo L, et al.: The explosion of non-nano iron dust suspension in the 20-l spherical bomb. Journal of Loss Prevention in the Process Industries. 2021; 71: 104447. Publisher Full Text
42. Cashdollar KL: Overview of dust explosibility characteristics. Journal of Loss Prevention in the Process Industries. 2000; 13(3–5): 183–199. Publisher Full Text
43. Clouthier M, Taveau JR, Dastidar AG, et al.: Iron and aluminum powder explosibility in 20-L and 1- m3 chambers. J Loss Prev Process Ind. 2019; 62: 103927. Publisher Full Text
44. Guo Y, Ren K, Wei A, et al.: Iron dust explosion characteristics with small amount of nano-sized Fe2O3 and Fe3O4 particles. Fuel. 2022; 324: 124786. Publisher Full Text
45. Semenova A, Angerler T, Hohenberger M, et al.: Experimental investigation of direct reduced iron dust explosion characteristics in the 20 L sphere. Proceedings of the 15th International Symposium on Hazards, Prevention and Mitigation of Industrial Explosions (ISHPMIE 2024). 2024.
46. Taveau JR, Lemkowitz SM, Hochgreb S, et al.: Metal dusts explosion hazards and protection. CHEMICAL ENGINEERING. 2019; 77.
47. Wu H-C, Chang R-C, Hsiao H-C: Research of minimum ignition energy for nano Titanium powder and nano Iron powder. Journal of Loss Prevention in the Process Industries. 2009; 22(1): 21–24. Publisher Full Text
48. Meng X, Wang Z, Zhang Y, et al.: Experimental study of explosion overpressure and flame propagation of micro-sized and nanosized iron powder. Process Safety Progress. 2023; 42(1): 116–125. Publisher Full Text
49. Myers TJ, Ibarreta AF, Marr K: Assessing the Hazard of Marginally Explosible Dusts. 9th Global Congress on Process Safety. 2013.
50. Addo A, Portarapillo M, Di Benedetto A, et al.: Investigation of marginally explosible dusts. J Loss Prev Process Ind. 2024; 87: 105246. Publisher Full Text
51. Proust C, Accorsi A, Dupont L: Measuring the violence of dust explosions with the “20l sphere” and with the standard “ISO 1m3 vessel”. Journal of Loss Prevention in the Process Industries. 2007; 20(4–6): 599–606. Publisher Full Text
52. van Rooij NE : Development of iron powder boilers for industry. TU Eindhoven. PhD thesis. 2025.
53. European Committee for Standardization (CEN): Determination of explosion characteristics of dust clouds: Part 2: Determination of the maximum rate of explosion pressure rise (dp/dt) max of dust clouds. 2011, EN 14034–2.
54. European Committee for Standardization (CEN): Determination of explosion characteristics of dust clouds: Part 1: Determination of the maximum explosion pressure (Pmax) of dust clouds. 2011, EN 14034–1.
55. European Committee for Standardization (CEN): Determination of explosion characteristics of dust clouds: Part 3: Determination of the lower explosion limit (LEL) of dust clouds. 2011, EN 14034–3.
56. European Committee for Standardization (CEN): Determination of explosion characteristics of dust clouds: Part 4: Determination of the limiting oxygen concentration (LOC) of dust clouds. 2011, EN 14034–4.
57. International Organization for Standardization (ISO): ISO/IEC 80079–20-2: Explosive Atmospheres – Part 20–2: Material Characteristics — Combustible Dusts Test Methods.2016.
58. American Society for Testing and Materials (ASTM): ASTM E1226–19 – Standard Test Method for Explosibility of Dust Clouds. 2019.
59. VDI: VDI 2263 Blatt 1:2022–02 – Test methods for the determination of the safety characteristics of dusts. 2022.
60. VDI: VDI 2263 – Staubbrände und Staubexplosionen – Gefahren – Beurteilung – Schutzmaßnahmen. 2018.
61. Choisez L, van Ende M-A , Bruyr Z, et al.: Influence of impurities on the use of Fe-based powder as sustainable fuel. Philos Trans A Math Phys Eng Sci. 2024; 382(2284): 20230236. Publisher Full Text
62. Choisez L, van Rooij NE , Hessels CJ, et al.: Phase transformations and microstructure evolution during combustion of iron powder. Acta Mater. 2022; 239: 118261. Publisher Full Text
63. Cesana C, Siwek R: “Manual for 20-l-Apparatus,” REMBE Research+Technology Center GmbH/Cesana AG, 3/27/2025.
64. Spitzer SH, Pandal A, Rodríguez MV, et al.: Investigations on different distribution systems for dusts inside the 20L-sphere. Journal of Loss Prevention in the Process Industries. 2024; 92: 105492. Publisher Full Text
65. Dufaud O, Janovsky B, Karl J, et al.: From Dust Dispersion to Dust Explosion. CHEMICAL ENGINEERING TRANSACTIONS. 2025; 116: 661–666.
66. International Organization for Standardization (ISO): ISO 18134-1:2015 – Solid biofuels – Determination of moisture content – Oven dry method – Part 1: Total moisture – Reference method.2015.
67. Chang S-C, Cheng Y-C, Zhang X-H, et al.: Effects of moisture content on explosion characteristics of incense dust in incense factory. Journal of Thermal Analysis and Calorimetry. 2022; 147(4): 2885–2892. Publisher Full Text
68. Santamaría-Herrera A, Hoyuelos FJ, Casado-Marcos C: Characterization of the explosiveness of wood dust. Process Safety and Environmental Protection. 2023; 169: 252–259. Publisher Full Text
69. International Organization for Standardization (ISO): ISO 13320:2020 – Particle size analysis – Laser diffraction methods.2020.
70. ÖNORM EN: 15415–1: Solid recovered fuels – Determination of particle size distribution – Part 1: Screen method for small dimension particles.
71. Whiting JG, Garboczi EJ, Tondare VN, et al.: A comparison of particle size distribution and morphology data acquired using lab-based and commercially available techniques: Application to stainless steel powder. Powder Technology. 2022; 396: 648–662. Publisher Full Text
72. Cen L, Qian Y, Lu X: Surface morphology effects on ignition temperature of single micron-sized Iron particles. Combustion and Flame. 2025; 277: 114216. Publisher Full Text
73. Janovsky B, Spitzer SH, Karl J, et al.: Simplifying standards, opening restrictions part II: The significance of the ignition source on the safety characteristics of dusts in the 20 L-sphere. Journal of Loss Prevention in the Process Industries. 2025; 94: 105524. Publisher Full Text
74. Spitzer SH: Influence of the ignition source on the safety characteristics of hybrid dust–gas mixtures. 2023.
75. Liu X, Hashimoto N, Xia Y, et al.: Experimental investigation of oxygen concentration effects on spherical laminar flame propagation of micron-sized iron particle clouds. Combustion and Flame. 2025; 277: 114234. Publisher Full Text
76. D’Amico M: Étude expérimentale et modélisation des explosions hybrides solides/solides: application au cas des mélanges de poussières graphite/métaux. 2016.
77. Semenova A, Spitzer SH: IRON ROUND ROBIN Semenova and Spitzer 2026.2026. Publisher Full Text

Footnotes

1 According to the European and International standards for the determination of safety characteristic of dusts, dust is defined as solid particles with a diameter below 500 μm.⁵³^–⁵⁷ The American standard applies for particles below 75 μm⁵⁸ and the German compulsive regulatory differentiates between dust layers (particles below 250 μm) and elevated dusts (particles below 63 μm).⁵⁹ However, the superordinate German regulatory again defines everything below 500 μm as a dust.⁶⁰

Comments

Publication information

Section

Research Article

Published

June 3, 2026

DOI

https://doi.org/10.12688/openreseurope.22839.2

Versions

June 3, 2026 (Published Manuscript Under Review 2.0)
February 18, 2026 (Published Manuscript Under Review 1.0)

Included in

Funding data

Slovak Research and Development Agency

Grant ID(s): APVV-21-0187
European Commission

Grant ID(s): 101186946
TKI-Chemie

Grant ID(s): CHEMIE.PGT2022.005

Metrics

Views

144

Downloads

Citations

See full metrics

Keywords

MeCRE, Safety Characteristics, Chemical Energy Carriers, Clean Energy Transition, Standardized Iron Fuel, Dust explosions

Comments

Aleksandra Semenova

Montanuniversitat Leoben, Leoben, Styria, Austria

Roles

Co-Author, Conceptualization, Data Curation, Resources, Visualization, Writing – Original Draft Preparation

0009-0005-2380-8082

aleksandra.semenova@unileoben.ac.at

Stefan H. Spitzer Sr

VSB-Technical University of Ostrava, Faculty of Safety Engineering, Centre of Excellence for Safety Research, Ostrava, Moravian-Silesian Region, 70800, Czech Republic

European Institute for Innovation - Technology, Schwaebisch-Gmuend, Baden-Wuerttemberg, 73525, Germany

3.7, Physikalisch-Technische Bundesanstalt, Brunswick, Lower Saxony, 38116, Germany

Roles

Corresponding Author, Co-Author, Conceptualization, Data Curation, Formal Analysis, Funding Acquisition, Investigation, Validation, Writing – Review & Editing

0000-0002-6851-9963

s.spitzer@eifi-tech.eu

Barbora Dvorakova

VSB-Technical University of Ostrava, Faculty of Safety Engineering, Centre of Excellence for Safety Research, Ostrava, Moravian-Silesian Region, 70800, Czech Republic

Roles

Co-Author, Data Curation, Funding Acquisition, Project Administration, Supervision

barbora.dvorakova@vsb.cz

Enis Askar

2, BAM Federal Institute for Materials Research and Testing, Berlin, Berlin, 12205, Germany

Roles

Co-Author, Investigation, Writing – Review & Editing

enis.askar@bam.de

Patrick Funnemann

2, BAM Federal Institute for Materials Research and Testing, Berlin, Berlin, 12205, Germany

Roles

Co-Author, Investigation, Writing – Review & Editing

0009-0007-5093-6092

patrick.funnemann@bam.de

Roy T.E. Hermanns

Technische Universiteit Eindhoven, Eindhoven, North Brabant, The Netherlands

Roles

Co-Author, Resources, Writing – Review & Editing

0000-0002-4907-7266

r.t.e.hermanns@tue.nl

Vojtech Jankuj

VSB-Technical University of Ostrava, Faculty of Safety Engineering, Centre of Excellence for Safety Research, Ostrava, Moravian-Silesian Region, 70800, Czech Republic

Roles

Co-Author, Investigation, Validation, Writing – Review & Editing

0000-0002-4199-406X

vojtech.jankuj@vsb.cz

Alexis Vignes

Institut National de l'Environnement Industriel et des Risques, Verneuil-en-Halatte, Nord-Pas-de-Calais-Picardie, France

Roles

Co-Author, Investigation, Writing – Review & Editing

0000-0003-0737-7911

alexis.vignes@ineris.fr

Richard Kuracina

Institute of Integral Safety, Slovak University of Technology in Bratislava, Trnava, Trnava, 917 08, Slovakia

Roles

Co-Author, Investigation, Writing – Review & Editing

0000-0003-1468-0820

richard.kuracina@stuba.sk

Zuzana Szabova

Institute of Integral Safety, Slovak University of Technology in Bratislava, Trnava, Trnava, 917 08, Slovakia

Roles

Co-Author, Investigation, Writing – Review & Editing

0000-0002-7886-1623

zuzana.szabova@stuba.sk

Frederik Norman

Adinex N.V., Noorderwijk-Herentals, B-2200, Belgium

Roles

Co-Author, Investigation, Writing – Review & Editing

frederik.norman@adinex.be

Dieter Gabel

Otto von Guericke Universitat Magdeburg Fakultat fur Verfahrens- und Systemtechnik, Magdeburg, Saxony-Anhalt, 39106, Germany

Roles

Co-Author, Investigation, Writing – Review & Editing

dieter.gabel@ovgu.de

Enrico Danzi

VSB-Technical University of Ostrava, Faculty of Safety Engineering, Centre of Excellence for Safety Research, Ostrava, Moravian-Silesian Region, 70800, Czech Republic

Polytechnic of Turin Department of Applied Science and Technology, Turin, Piedmont, 10129, Italy

Roles

Co-Author, Investigation, Writing – Review & Editing

enrico.danzi@vsb.cz

Ekaterina Makarova

Pometon S.p.A., Veneto, 30030, Italy

Roles

Co-Author, Resources

ekaterina.makarova@pometon.com

Ivan Lorenzon

Pometon S.p.A., Veneto, 30030, Italy

Roles

Co-Author, Resources

ivan.lorenzon@pometon.com

Adrian Toman

Experimental Mine "Barbara", Central Mining Institute - National Research Institute, Mikolow, Silesian, 43-190, Poland

Roles

Co-Author, Investigation, Writing – Review & Editing

0000-0002-2879-9702

atoman@gig.eu

Wojciech Adamus

Experimental Mine "Barbara", Central Mining Institute - National Research Institute, Mikolow, Silesian, 43-190, Poland

Roles

Co-Author, Investigation, Writing – Review & Editing

0000-0001-8948-4889

wadamus@gig.eu

Conrad Hessels

Technische Universiteit Eindhoven, Eindhoven, North Brabant, The Netherlands

Roles

Co-Author, Investigation, Writing – Review & Editing

0000-0002-7619-8418

c.j.m.hessels@tue.nl

Downloads over the last 12 months:

Cited by

via Crossref

View citing articles

[1] 1. Bergthorson JM, Goroshin S, Soo MJ, et al.: Direct combustion of recyclable metal fuels for zero-carbon heat and power. Applied Energy. 2015; 160: 368–382. Publisher Full Text

[2] 2. Metalot: It’s time for Circular Energy: It’s time for Iron Power—Iron Power Program: Strategic Rationale.2025.

[3] 3. Neumann J, Da Rocha RC, Debiagi, et al.: Techno-economic assessment of long-distance supply chains of energy carriers: Comparing hydrogen and iron for carbon-free electricity generation. Appl Energy Combust Sci. 2023; 14: 100128. Publisher Full Text

[4] 4. Church C, Crawford A: Minerals and the Metals for the Energy Transition: Exploring the Conflict Implications for Mineral-Rich :Fragile States; 2020.

[5] 5. Bock S, Pauritsch M, Lux S, et al.: Natural iron ores for large-scale thermochemical hydrogen and energy storage. Energy Conversion and Management. 2022; 267: 115834. Publisher Full Text

[6] 6. Voitic G, Hacker V: Recent advancements in chemical looping water splitting for the production of hydrogen. RSC Advances. 2016; 6(100): 98267–98296. Publisher Full Text

[7] 7. Quitzow R, Zabanova Y: The Geopolitics of Hydrogen: Volume 1: European Strategies in Global Perspective. Cham: Springer Nature Switzerland; Imprint Springer; 2024.

[8] 8. Hafner M, Tagliapietra S: The geopolitics of the global energy transition. Cham: Springer Open; 2020.

[9] 9. Hydrogen Research Europe: Advancing Hydrogen Technologies – A roadmap for pioneering research projects. Hydrogen Research Europe. November 2025.

[10] 10. Geurts B, Beijer A: World’s first in fossil-free industrial heat – RIFT and Kingspan Unidek sign historic first commercial contract for Iron Fuel Technology. 2025.https://www.kingspangroup.com/en/news-insights/rift-and-kingspan-unidek-sign-historic-first-commercial-contract/

[11] 11. Boretti A, Castelletto S: Metal fuels for renewable energy storage and circular economy. Next Energy. 2025; 8: 100371. Publisher Full Text

[12] 12. Eckhoff RK: Dust explosions in the process industries: Identification, assessment and control of dust hazards. Gulf Professional Publishing; 2003.

[13] 13. Manilevich FD, Pirskyy Y, Kutsyi AV, et al.: Aluminium as an efficient energy storage substance for the catalysed generation of hydrogen from water. J Energy Storage. 2024; 96: 112748. Publisher Full Text

[14] 14. Kim W, Saeki R, Ueno Y, et al.: Effect of particle size on the minimum ignition energy of aluminum powders. Powder Technology. 2023; 415: 118190. Publisher Full Text

[15] 15. World Nuclear Association: Heat Values of Various Fuels. 2026.https://world-nuclear.org/

[16] 16. Kecojevic V, Vukotic I, Komljenovic D: Production, consumption and cost of energy for surface mining of bituminous coal. Min Eng. 2014; 66: 51–57.

[17] 17. Norman F, Berghmans J, Verplaetsen F: The Minimum Ignition Energy of Coal Dust in an Oxygen Enriched Atmosphere. Chem Eng Trans. 2013; 31.

[18] 18. Pedišius N, Praspaliauskas M, Pedišius J, et al.: Analysis of Wood Chip Characteristics for Energy Production in Lithuania. Energies. 2021; 14(13): 3931. Publisher Full Text

[19] 19. CROHMIQ: Minimum Ignition Energy Data by Product. 2026.https://crohmiq.com/minimum-ignition-energy-data-by-product/

[20] 20. DECHEMA: CHEMSAFE – Explosion Characteristics of Substances and Mixtures. 2026.https://dechema.de/en/chemsafe.html

[21] 21. Pan C, Wang X, Li G, et al.: Influences of a square and circular vent on gasoline vapor explosions and its propagation: Comparative experimental study. Case Studies in Thermal Engineering. 2021; 27: 101225. Publisher Full Text

[22] 22. Ogle RA: Dust Explosion Dynamics. Butterworth-Heinemann, Elsevier; 2016.

[23] 23. Bidabadi M, Mafi M: Time variation of combustion temperature and burning time of a single iron particle. International Journal of Thermal Sciences. 2013; 65: 136–147. Publisher Full Text

[24] 24. Panahi A, Chang D, Schiemann M, et al.: Combustion behavior of single iron particles-part I: An experimental study in a drop-tube furnace under high heating rates and high temperatures. Appl Energy Combust Sci. 2023; 13: 100097. Publisher Full Text

[25] 25. Mi X, Fujinawa A, Bergthorson JM: A quantitative analysis of the ignition characteristics of fine iron particles. Combustion and Flame. 2022; 240: 112011. Publisher Full Text

[26] 26. Fujinawa A, Thijs LC, Jean-Philyppe J, et al.: Combustion behavior of single iron particles, Part II: A theoretical analysis based on a zero-dimensional model. Applications in Energy and Combustion Science. 2023; 14: 100145. Publisher Full Text

[27] 27. Ning D, Shoshin Y, van Oijen JA , et al.: Burn time and combustion regime of laser-ignited single iron particle. Combustion and Flame. 2021; 230: 111424. Publisher Full Text

[28] 28. Ning D, Hazenberg T, Shoshin Y, et al.: Experimental and theoretical study of single iron particle combustion under low-oxygen dilution conditions. Fuel. 2024; 357: 129718. Publisher Full Text

[29] 29. Leshchevich VV, Penyazkov OG, Fedorov AV, et al.: Conditions and delay time of ignition of iron microparticles in oxygen. Journal of Engineering Physics and Thermophysics. 2012; 85: 148–154. Publisher Full Text

[30] 30. Vance FH, Scholtissek A, Nicolai H, et al.: Flame propagation modes for iron particle clusters in air — Part I: Transition from continuous to discrete propagation mode under weak convection effects. Combustion and Flame. 2024; 260: 113265. Publisher Full Text

[31] 31. Vance FH, Scholtissek A, Nicolai H, et al.: Flame propagation modes for iron particle clusters in air, Part II: Transition from continuous to discrete propagation mode under strong convection effects. Combustion and Flame. 2024; 265: 113199. Publisher Full Text

[32] 32. Fujinawa A, Mi X: Numerical studies on the propagation of iron dust flames in confinement. Proceedings of the Combustion Institute. 2024; 40(1–4): 105754. Publisher Full Text

[33] 33. Goroshin S, Tang F-D, Higgins AJ, et al.: Laminar dust flames in a reduced-gravity environment. Acta Astronautica. 2011; 68(7–8): 656–666. Publisher Full Text

[34] 34. Tang F-D, Goroshin S, Higgins A, et al.: Flame propagation and quenching in iron dust clouds. Proceedings of the Combustion Institute. 2009; 32(2): 1905–1912. Publisher Full Text

[35] 35. Tang F-D, Goroshin S, Higgins AJ: Modes of particle combustion in iron dust flames. Proceedings of the Combustion Institute. 2011; 33(2): 1975–1982. Publisher Full Text

[36] 36. Hazenberg T, Braig D, Fedoryk MA, et al.: Analyzing iron dust bunsen flames using numerical simulations. Fuel. 2026; 403: 136094. Publisher Full Text

[37] 37. Fedoryk M, Stelzner B, Harth S, et al.: Experimental investigation of the laminar burning velocity of iron-air flames in a tube burner. Applications in Energy and Combustion Science. 2023; 13: 100111. Publisher Full Text

[38] 38. van Gool C , Hazenberg T, van Oijen JA , et al.: Numerical determination of iron dust laminar flame speeds with the counter-flow twin-flame technique. Combust Flame. 2024; 266: 113524. Publisher Full Text

[39] 39. Spitzer sen SH, Stevens NC, Krietsch A, et al.: A first step towards a standardized iron fuel: Safety characteristics of several pure iron powders. Fuel. 2025; 400: 135824. Publisher Full Text

[40] 40. Danzi E, Marmo L: Dust explosion risk in metal workings. Journal of Loss Prevention in the Process Industries. 2019; 61: 195–205. Publisher Full Text

[41] 41. Danzi E, Pio G, Marmo L, et al.: The explosion of non-nano iron dust suspension in the 20-l spherical bomb. Journal of Loss Prevention in the Process Industries. 2021; 71: 104447. Publisher Full Text

[42] 42. Cashdollar KL: Overview of dust explosibility characteristics. Journal of Loss Prevention in the Process Industries. 2000; 13(3–5): 183–199. Publisher Full Text

[43] 43. Clouthier M, Taveau JR, Dastidar AG, et al.: Iron and aluminum powder explosibility in 20-L and 1- m3 chambers. J Loss Prev Process Ind. 2019; 62: 103927. Publisher Full Text

[44] 44. Guo Y, Ren K, Wei A, et al.: Iron dust explosion characteristics with small amount of nano-sized Fe2O3 and Fe3O4 particles. Fuel. 2022; 324: 124786. Publisher Full Text

[45] 45. Semenova A, Angerler T, Hohenberger M, et al.: Experimental investigation of direct reduced iron dust explosion characteristics in the 20 L sphere. Proceedings of the 15th International Symposium on Hazards, Prevention and Mitigation of Industrial Explosions (ISHPMIE 2024). 2024.

[46] 46. Taveau JR, Lemkowitz SM, Hochgreb S, et al.: Metal dusts explosion hazards and protection. CHEMICAL ENGINEERING. 2019; 77.

[47] 47. Wu H-C, Chang R-C, Hsiao H-C: Research of minimum ignition energy for nano Titanium powder and nano Iron powder. Journal of Loss Prevention in the Process Industries. 2009; 22(1): 21–24. Publisher Full Text

[48] 48. Meng X, Wang Z, Zhang Y, et al.: Experimental study of explosion overpressure and flame propagation of micro-sized and nanosized iron powder. Process Safety Progress. 2023; 42(1): 116–125. Publisher Full Text

[49] 49. Myers TJ, Ibarreta AF, Marr K: Assessing the Hazard of Marginally Explosible Dusts. 9th Global Congress on Process Safety. 2013.

[50] 50. Addo A, Portarapillo M, Di Benedetto A, et al.: Investigation of marginally explosible dusts. J Loss Prev Process Ind. 2024; 87: 105246. Publisher Full Text

[51] 51. Proust C, Accorsi A, Dupont L: Measuring the violence of dust explosions with the “20l sphere” and with the standard “ISO 1m3 vessel”. Journal of Loss Prevention in the Process Industries. 2007; 20(4–6): 599–606. Publisher Full Text

[52] 52. van Rooij NE : Development of iron powder boilers for industry. TU Eindhoven. PhD thesis. 2025.

[53] 53. European Committee for Standardization (CEN): Determination of explosion characteristics of dust clouds: Part 2: Determination of the maximum rate of explosion pressure rise (dp/dt) max of dust clouds. 2011, EN 14034–2.

[54] 54. European Committee for Standardization (CEN): Determination of explosion characteristics of dust clouds: Part 1: Determination of the maximum explosion pressure (Pmax) of dust clouds. 2011, EN 14034–1.

[55] 55. European Committee for Standardization (CEN): Determination of explosion characteristics of dust clouds: Part 3: Determination of the lower explosion limit (LEL) of dust clouds. 2011, EN 14034–3.

[56] 56. European Committee for Standardization (CEN): Determination of explosion characteristics of dust clouds: Part 4: Determination of the limiting oxygen concentration (LOC) of dust clouds. 2011, EN 14034–4.

[57] 57. International Organization for Standardization (ISO): ISO/IEC 80079–20-2: Explosive Atmospheres – Part 20–2: Material Characteristics — Combustible Dusts Test Methods.2016.

[58] 58. American Society for Testing and Materials (ASTM): ASTM E1226–19 – Standard Test Method for Explosibility of Dust Clouds. 2019.

[59] 59. VDI: VDI 2263 Blatt 1:2022–02 – Test methods for the determination of the safety characteristics of dusts. 2022.

[60] 60. VDI: VDI 2263 – Staubbrände und Staubexplosionen – Gefahren – Beurteilung – Schutzmaßnahmen. 2018.

[61] 61. Choisez L, van Ende M-A , Bruyr Z, et al.: Influence of impurities on the use of Fe-based powder as sustainable fuel. Philos Trans A Math Phys Eng Sci. 2024; 382(2284): 20230236. Publisher Full Text

[62] 62. Choisez L, van Rooij NE , Hessels CJ, et al.: Phase transformations and microstructure evolution during combustion of iron powder. Acta Mater. 2022; 239: 118261. Publisher Full Text

[63] 63. Cesana C, Siwek R: “Manual for 20-l-Apparatus,” REMBE Research+Technology Center GmbH/Cesana AG, 3/27/2025.

[64] 64. Spitzer SH, Pandal A, Rodríguez MV, et al.: Investigations on different distribution systems for dusts inside the 20L-sphere. Journal of Loss Prevention in the Process Industries. 2024; 92: 105492. Publisher Full Text

[65] 65. Dufaud O, Janovsky B, Karl J, et al.: From Dust Dispersion to Dust Explosion. CHEMICAL ENGINEERING TRANSACTIONS. 2025; 116: 661–666.

[66] 66. International Organization for Standardization (ISO): ISO 18134-1:2015 – Solid biofuels – Determination of moisture content – Oven dry method – Part 1: Total moisture – Reference method.2015.

[67] 67. Chang S-C, Cheng Y-C, Zhang X-H, et al.: Effects of moisture content on explosion characteristics of incense dust in incense factory. Journal of Thermal Analysis and Calorimetry. 2022; 147(4): 2885–2892. Publisher Full Text

[68] 68. Santamaría-Herrera A, Hoyuelos FJ, Casado-Marcos C: Characterization of the explosiveness of wood dust. Process Safety and Environmental Protection. 2023; 169: 252–259. Publisher Full Text

[69] 69. International Organization for Standardization (ISO): ISO 13320:2020 – Particle size analysis – Laser diffraction methods.2020.

[70] 70. ÖNORM EN: 15415–1: Solid recovered fuels – Determination of particle size distribution – Part 1: Screen method for small dimension particles.

[71] 71. Whiting JG, Garboczi EJ, Tondare VN, et al.: A comparison of particle size distribution and morphology data acquired using lab-based and commercially available techniques: Application to stainless steel powder. Powder Technology. 2022; 396: 648–662. Publisher Full Text

[72] 72. Cen L, Qian Y, Lu X: Surface morphology effects on ignition temperature of single micron-sized Iron particles. Combustion and Flame. 2025; 277: 114216. Publisher Full Text

[73] 73. Janovsky B, Spitzer SH, Karl J, et al.: Simplifying standards, opening restrictions part II: The significance of the ignition source on the safety characteristics of dusts in the 20 L-sphere. Journal of Loss Prevention in the Process Industries. 2025; 94: 105524. Publisher Full Text

[74] 74. Spitzer SH: Influence of the ignition source on the safety characteristics of hybrid dust–gas mixtures. 2023.

[75] 75. Liu X, Hashimoto N, Xia Y, et al.: Experimental investigation of oxygen concentration effects on spherical laminar flame propagation of micron-sized iron particle clouds. Combustion and Flame. 2025; 277: 114234. Publisher Full Text

[76] 76. D’Amico M: Étude expérimentale et modélisation des explosions hybrides solides/solides: application au cas des mélanges de poussières graphite/métaux. 2016.

[77] 77. Semenova A, Spitzer SH: IRON ROUND ROBIN Semenova and Spitzer 2026.2026. Publisher Full Text

Towards standardized safety protocols for iron-based energy carriers: International alignment through round robin testing on safety characteristics

List of Authors

Peer review

Publication information

Included in

Funding data

Metrics

Keywords

Comments

Abstract

Amendment

Comments

Publication information

Included in

Funding data

Metrics

Keywords

Comments

Authors & roles

Aleksandra Semenova

Stefan H. Spitzer Sr

Barbora Dvorakova

Enis Askar

Patrick Funnemann

Roy T.E. Hermanns

Vojtech Jankuj

Alexis Vignes

Richard Kuracina

Zuzana Szabova

Frederik Norman

Dieter Gabel

Enrico Danzi

Ekaterina Makarova

Ivan Lorenzon

Adrian Toman

Wojciech Adamus

Conrad Hessels

Downloads over the last 12 months:

Cited by

via Crossref