Fire and explosion properties of nanopowders - HSE

Health and Safety Executive

Fire and explosion properties of nanopowders Prepared by the Health and Safety Laboratory for the Health and Safety Executive 2010

RR782 Research Report

Health and Safety Executive

Fire and explosion properties of nanopowders P Holbrow, M Wall, E Sanderson, D Bennett, W Rattigan, R Bettis & D Gregory Health and Safety Laboratory Harpur Hill Buxton Derbyshire SK17 9JN

Nanotechnology is a rapidly expanding technology in which existing and novel materials are engineered at the nanoscale, typically in the range of 1 to 100 nanometres. Engineered nanomaterials include uniquely manufactured products with unique shapes and enhanced physical and chemical properties, compared with conventional materials of the same composition. There is currently little available information on the explosion risks of these materials. The Health and Safety Executive therefore commissioned this project to investigate the potential fire and explosion hazards associated with nanopowders. Test equipment and procedures were developed to assess the key properties of a selected number of nanopowders. A specialised 2 litre test vessel was developed to determine the explosion characteristics and modified standard test apparatus was used to measure the minimum ignition energy of nanopowders. Resistivity and electrostatic charging characteristics were assessed using specially designed test apparatus. Key information including KSt, Pmax and MIE values were obtained for a range of metal and carbon nanopowders. Generally, the explosibility (maximum explosion pressure, rates of pressure rise and equivalent KSt) of nanopowders were found to be broadly similar to conventional micron-scale powders. However, the minimum ignition energies of some nanopowders were found to be lower than the equivalent material at micron-scale. It was demonstrated that with increasing relative humidity the resistivity of most nanopowders decreases. There was also a tendency for nanopowders to have higher resistivity values than conventional micron-scale powders. All the powders produced electrostatic charge. Generally, the charge developed by nanopowders was comparable with the micron-scale powders. This report and the work it describes were funded by the Health and Safety Executive (HSE). Its contents, including any opinions and/or conclusions expressed, are those of the authors alone and do not necessarily reflect HSE policy.

HSE Books

© Crown copyright 2010 First published 2010 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means (electronic, mechanical, photocopying, recording or otherwise) without the prior written permission of the copyright owner. Applications for reproduction should be made in writing to: Licensing Division, Her Majesty’s Stationery Office, St Clements House, 2-16 Colegate, Norwich NR3 1BQ or by e-mail to [email protected]

ii

CONTENTS 1

INTRODUCTION......................................................................................... 1

2

LEGAL BACKGROUND............................................................................. 2

3 EXPLOSION SEVERITY............................................................................. 3 3.1 Standard test ........................................................................................... 3 3.2 Test vessel .............................................................................................. 4 3.3 Control and Instrumentation .................................................................... 7 3.4 Test procedure ........................................................................................ 8 4 MINIMUM IGNITION ENERGY ................................................................. 10 4.1 Test apparatus....................................................................................... 10 5 FIRE HAZARDS ....................................................................................... 13 5.1 Fire Hazards in nanopowders................................................................ 13 5.2 Practical Assessment of Fire Properties ................................................ 13 6 ELECTROSTATIC PROPERTIES ............................................................ 15 6.1 Electrostatic charge ............................................................................... 15 6.2 Charging test ......................................................................................... 16 6.3 Resistivity test........................................................................................ 17 7

TEST MATERIALS ................................................................................... 19

8 RESULTS ................................................................................................. 22 8.1 Commissioning tests – explosion severity ............................................. 22 8.2 Explosion severity of nanopowders ....................................................... 28 8.3 Commissioning MIE test results ............................................................ 34 8.4 MIE test results – nanopowders ............................................................ 35 8.5 Electrostatics ......................................................................................... 37 9 DISCUSSION............................................................................................ 54 9.1 Explosion test Equipment ...................................................................... 54 9.2 Explosion severity and MIE ................................................................... 54 9.3 Electrostatic issues ................................................................................ 58 10

CONCLUSIONS .................................................................................... 61

11

APPENDICES ....................................................................................... 63

12

REFERENCES ...................................................................................... 68

iii

iv

EXECUTIVE SUMMARY

Objectives Nanotechnology is a rapidly expanding technology in which existing and novel materials are engineered at the nanoscale. However, there is currently little available information on the fire and explosion risks of nanoparticles. The objective of this project is to characterize the fire and explosion properties of a range of commercially available nanopowders.

Main Findings 1. A unique facility has been developed specifically to allow the safe handling and testing of nanopowders. The equipment comprises: i)

2 litre explosion test apparatus for measuring the rate of pressure rise and the maximum explosion pressure.

ii)

Minimum ignition energy test apparatus. The minimum ignition energy of nanopowders was measured using a modified Kuhner MIKE3 test apparatus modified to allow nanopowders to be safely handled within the test apparatus.

iii)

Sealed systems that allow the safe handling of oxidized and non-oxidized nanopowders under inert atmospheres.

2. Explosion properties of a range of nanopowders have been characterised including aluminium, iron, zinc, copper and several carbon nanomaterials. Generally, the explosibility (maximum explosion pressure, rates of pressure rise and equivalent KSt) of nanopowders are broadly similar to conventional micron-scale powders. The minimum ignition energies of some nanopowders have been lower than the equivalent material at micron-scale. This indicates that the nanopowders may be more susceptible to ignition but once ignited the explosion violence is no more severe than micron-scale powders. 3. Electrostatic properties of nanopowders have been considered. Dust layer resistivity measurements and charge tests have been made on a range of nanopowders.

Recommendations To build on the data already obtained, and develop further expertise and knowledge, further nanopowders should be obtained and characterised.

v

vi

1

INTRODUCTION

Nanotechnology is a rapidly expanding technology in which existing and novel materials are engineered at the nanoscale. Engineered nanomaterials include uniquely manufactured products with unique shapes and enhanced physical and chemical properties, compared with conventional materials of the same composition. The Health and Safety Executive has commissioned a project to investigate the potential hazards associated with nanopowders. The general method of approach is to use standard, and modified, fire and explosion characterization methods to assess the key properties of a selected number of nanopowders. As part of this project, a review (HSL report number XS/08/119) has been carried out to update a review done by HSL in 2003 and to ensure that all relevant new material in this fast changing area is incorporated. There is currently little available information on the explosion risks of nanoparticles. The little data that does exist is contradictory and suggests the explosion violence is less than that of larger particles, which goes against the expectation that smaller particles burn faster than larger ones due to the increased surface area. Fine powders are known to be an explosion risk, particularly organic and metallic powders. For fine particulates, there is standard test equipment available to determine the explosion and ignition characteristics for powders, but these typically require kilogram quantities of powder. Also, the powder is dispersed in the standard apparatus using compressed air. With nanopowders, their large surface to volume ratio means that many are spontaneously flammable on contact with air, or surface oxidation alters their properties. Hence equipment that avoids oxidation until the point of ignition has been developed. This project report describes the design, operation of test apparatus and results obtained relating to the explosion properties of nanopowders. Key information including explosion severity, minimum ignition energy and electrostatic properties is reported.

1

2

LEGAL BACKGROUND

The major legal regulations that could apply to nanopowders include the following: DSEAR Regulations It is the duty of employers to comply with the Dangerous Substances and Explosive Atmosphere Regulations 2002 (DSEAR) which seek to eliminate, reduce and control the fire and explosion risks from dangerous substances. Flammable nanopowders capable of fueling a dust explosion fall within the definition of a dangerous substance. Regulation 5 of these regulations requires that, where a dangerous substance is present at a workplace, the employer should make a suitable and sufficient assessment of the risks to their employees. Those risks should, where possible, be eliminated. Where it is not reasonably practicable to do this they should be reduced and controlled (Regulation 6). Measuring the fire explosion characteristics of nanopowders is an essential step in assessing the risk. CHIP Regulations The CHIP (Chemical Hazard Information and Packaging for Supply) Regulations 2009 have implications for nanopowders. Dangerous substances or preparations cannot be supplied unless they have been classified in accordance with the regulations. Classification requires a knowledge of the toxic and fire and explosion properties of the substance. In addition, suppliers are required to provide information about the hazards of the substances and to package and label them for safe transport. REACH Regulations REACH is a new European Union regulation concerning the Registration, Evaluation, Authorisation and restriction of CHemicals. It came into force on 1st June 2007 and replaces a number of European Directives and Regulations with a single system. A major part of REACH is the requirement for manufacturers or importers of substances to register them with a central European Chemicals Agency (ECHA). A registration package will be supported by a standard set of data on that substance. The amount of data required is proportionate to the amount of substance manufactured or supplied. The data requirement includes the fire and explosion properties of the substance.

2

3 3.1

EXPLOSION SEVERITY

STANDARD TEST

The procedure for measuring the explosion severity of dust/air mixtures is described in a European standard available as BS EN 14034-1 (2004) and BS EN 14034-2 (2006). The standard test vessel for these determinations is the 1 m3 vessel. The peak maximum explosion pressure, Pmax, and the peak maximum rate of pressure rise, (dP/dt)max, are measured in this standard test procedure. The Pmax and (dP/dt)max are the highest values generated in an enclosed dust explosion. These characteristics are measured in a standard test at the optimum dust concentration and are obtained by testing the dust over a wide range of dust concentrations. The two peaks normally occur at different dust concentrations. The (dP/dt)max is used to calculate a dust specific explosibility characteristic called the KSt value. The KSt is given by: KSt = (dP/dt)maxV1/3

(2.1)

Where (dP/dt)max is the peak maximum rate of pressure rise (bar/s) and V is the total internal volume of the test vessel (m3). The units of KSt are bar m/s. Equation (2.1) is well known as the “cubic law” (Barton (2002)). The KSt is considered to be a constant for any dust, independent of vessel size and equation (2.1) acts as a simple scaling law. The KSt value is derived only from measurements in either a 20 litre sphere vessel or a 1 m3 vessel and if any other vessel is used to measure the KSt it must be calibrated against the 1 m3 or 20 litre standard test vessels. The ISO standard also allows the use of alternative vessels provided they give comparable results. The criterion for demonstrating conformity is given in the standard BS EN 14034-2 (2006); the standard has been prepared by CEN Technical Committee 305. This gives alternative test equipment procedures. It states that the maximum rate of pressure rise of dust clouds can be determined using alternative test equipment and/or test procedures. When using an alternative it shall be shown that at least for the following dusts that the method yields results within specific deviations. The dusts shall include at least two metal powders, two natural organic powders, two synthetic organic powders and two coal dusts. In the standard 1 m3 vessel the KSt value is equal to the maximum rate of pressure rise. The standard test apparatus typically requires kilogram quantities of powder and therefore a smaller scale test apparatus is required for nanopowders. Also, the powder is dispersed in the standard apparatus using compressed air. With nanopowders, their large surface to volume ratio means that many are spontaneously flammable on contact with air, or surface oxidation alters their properties. Hence equipment that avoids oxidation until the point of ignition is required. No commercial equipment is available to satisfy these requirements, so a specially designed test apparatus, to measure the explosion characteristics of nanopowders, has been developed in this project.

3

3.2

TEST VESSEL

An explosion test vessel with a volume of approximately 2 litres has been designed and manufactured specifically for testing nanopowders (Figure 1 - 3). The construction quality was specified to be similar to the existing standard 20 litre sphere apparatus at HSL but with specially treated internal surfaces. The vessel is constructed from stainless steel and designed for a maximum working pressure of 20 bar. It is fully certified and tested to British Standards and in accordance with the Pressure Systems Directive PD 5500: 2005. The shape of the internal test chamber is spherical since this will allow the dust to be dispersed as a homogeneous dust cloud and thus allow the burning material to expand uniformly towards the vessel wall. The internal volume of the spherical test chamber is 2 litres with an internal diameter of 156 mm. A twin-skinned wall to facilitate water-cooling as used in the 20 litre sphere apparatus was considered. This was rejected on the grounds that much of the vessel body will be congested with flanges and ports and the cooling would be ineffective. The vessel is therefore constructed with a single skin wall. The whole vessel assembly is capable of electrical isolation. Ports The test vessel incorporates a number of ports and flanged connections: 1. The main access into the vessel is via the top of the vessel. This access is sealed using a flanged plate that incorporates two holes to accept the ignition electrodes. The opening is sized to enable hand-access for cleaning and for installing the dispersion nozzle. 2. Two diametrically opposed flanges tapped with M14 threads for Kistler pressure transducers. 3. A port incorporating a glass viewing window with a diameter of 40 mm. 4. A dust injection port is located at the bottom of the vessel. 5. A pressure release valve and an evacuation port. Dust injection The dust is introduced into the vessel from an external dust container via a dust injection valve and solenoid system mounted on an intermediate flanged adapter at the base of the vessel and a dispersion nozzle. Dispersion nozzle A pepper-pot nozzle incorporating a rebound plate was selected for the test programme. The pepper-pot design with a circular deflector is designed to encourage a homogeneous dust distribution and is located into the base of the test chamber. Dust container A special dust container, designed for a working pressure of 20 bar, is attached to the dust injector. The container has a volume of approximately 0.06 litre. A ¼ inch BSP port is used for 4

the attachment of a pressure hose and pressure gauge. The container was designed to allow remote loading of nanopowder under an inert atmosphere within a glovebox. Ignition Ignition of the dust cloud is achieved using an ignition source attached to two central electrodes. There are a number of possible ignition sources including: chemical igniters, electric fuse heads, spark and hot wire. A voltage is applied to the ignition source from the control system at the preset time delay following dust injection into the vessel.

Figure 1: Vessel general arrangement

5

Figure 2: Test vessel

Figure 3: Test vessel

The test vessel was located within a fume cupboard, that incorporated a HEPA filter system and a sink fitted with waste filters for cleaning the contaminated components. Located outside the fume cupboard and connected via a pass-through port were: vacuum pump, high pressure air supply, control system and data acquisition system and a glovebox fitted with an integral HEPA 6

filter. To minimise operator contact with nanopowders, the materials were handled, weighed and loaded into the enclosed dust injection chamber within the glovebox (Figure 4). The glovebox also allowed the handling of nanopowders under an argon atmosphere, enabling powders to be transferred from the glovebox to the test apparatus under argon thus preventing premature oxidation of powders.

Figure 4: Glovebox 3.3

CONTROL AND INSTRUMENTATION

The control and instrumentation system is interfaced to a PC via custom control circuitry and commercial USB data acquisition and control hardware. The PC executes software providing control, data logging and analysis functions. The fume cupboard is fitted with safety interlocks to help ensure operator safety. The software has been developed using the National Instruments LabVIEW development environment. It is installed for use on the Microsoft Windows XP platform. In brief the test operates as follows. The sphere is sealed and evacuated. The powder under investigation is released into the sphere to form a cloud, driven by the in-rush of air from external pressure. The cloud is ignited, and the pressure rise in the sphere logged. This data is then analysed to determine the maximum rate of pressure rise and the peak pressure. The software controls the nanopowder release and ignition, logs the pressure data from the enclosure, and will analyse the data. Safety interlock status is monitored to prevent risk to personnel from unplanned attempts at powder release or ignition. The PC has two free USB 2.0 ports to allow interfacing to the USB data acquisition devices. The software provides appropriate interface signals via the interface hardware to control the apparatus. The software accepts appropriate interface signals to provide feedback for the control process and to log the required test data from two independent piezoelectric pressure sensors. The pressure transducers measure the pressure difference, not absolute pressure.

7

The software calculates dP/dt (the rate of pressure rise) and Pmax (maximum explosion overpressure) following each test. These are calculated for pressure sensor 1, for pressure sensor 2, and for the averaged readings of the two sensors. Kistler (type 701A) pressure transducers are used in the test apparatus and the details are shown in Table 1.

Laboratory instrument reference PT140

3.4

Channel

Table 1: Pressure transducers Serial number Transducer location

1 2

128026 128027

L.H. port R.H. port

Charge amplifier sensitivity 152.1 154.1

TEST PROCEDURE

The following is a summary of the general operational procedure. Sample preparation 1. Remove container with nanopowder from the storage cupboard. Place in a secondary transport container to protect the nanopowder container from accidental breakage. Transport to the test facility. 2. Remove the nanopowder container from the transport container a place in the glove box. The balance will be located in the glovebox. 3. Weigh the required mass of nanopowder by carefully scooping the powder and placing on a stainless steel dish on the balance. Close the nanopowder container. 4. Pour the nanopowder into the dust injection vessel and seal the vessel. Transfer the vessel to the fume cupboard, fit to the dust injection vessel and connect the supply lines. 5. Non-oxidised nanopowder is handled under an inert atmosphere. Firing 1. Prepare the ignition source. 2. If a chemical igniter is used, connect to the electrodes inside the vessel and check the circuit continuity using an approved ohmmeter tester. 3. If spark electrodes are used, ensure that the gap is set and the correct discharge energy is selected. 4. Replace the vessel lid and ensure a full complement of screws are fitted and are tightened. 5. Close the vessel outlet valve, open the evacuation valve and open the dust injection isolation valve.

8

6. Partially evacuate the explosion vessel to the appropriate pressure, typically 0.4 bara. The gas mixture and pressure in the explosion vessel shall be adjusted so that at ignition the atmosphere is nominally atmospheric. 7. Pressurise the dust injection vessel with compressed air. The pressure and volume of the dust injection vessel is matched to the level and vacuum in the explosion vessel such that the pressure at ignition is at nominal atmospheric pressure. 8. Close the fume cupboard sash and initiate the test. 9. Display the pressure traces and save the data to hard disk for analysis. 10. Open the outlet valve and release the pressure.

9

4 4.1

MINIMUM IGNITION ENERGY

TEST APPARATUS

The MIKE3 test apparatus, manufactured by Kuhner is normally used for the determination of minimum ignition energy (MIE) of micron-sized dusts (Figure 5). However, the risk assessment for handling nanopowders required improvements to the operating procedure for this apparatus. This equipment has therefore been modified to enable its use with nanopowders.

Figure 5: MIE test apparatus Dust dispersion chamber The dust dispersion chamber was modified to minimised operator contact with the powder and enable the safe handling, weighing of nanopowder and loading of the powder into the apparatus. This was achieved using a glovebox to weigh and transfer nanopowder to the MIKE3 apparatus. A two stage process was used to develop the modified dispersion chamber. Development Stage 1: An enclosed barrel shaped chamber was constructed to completely replace the Kuhner dispersion cup assembly (Figure 6).

10

Figure 6: Dispersion chamber This comprised a stainless steel chamber split into three components, lower middle and top barrels that were screwed together to form the assembled chamber. Two aluminium bursting discs were sandwiched between the barrels to form a central chamber that held the nanopowder. The assembly method was as follows: Powder, dispersion chamber and balance were placed in a glovebox. The lower and middle barrels were assembled with the lower bursting disc inserted between the barrels. Powder was then deposited into the middle chamber, the second bursting disc is fitted and the top barrel is screwed into position. The assembly was removed from the glovebox and transferred and fitted to the MIKE3 apparatus. The compressed air line was connected to the lower chamber. This provided the blast of compressed air that, during the test, ruptured the bursting discs and dispersed the nanopowder into the glass tube. Following commissioning tests on the first design it became clear that the efficacy of the system was not acceptable in that it was not possible to obtain the established MIE of a standard test powder. This was likely to be the result of erratic rupturing of the bursting discs and the increased distance of the initial formation of the dust cloud to the ignition source. Development Stage 2: A second dispersion system was therefore developed utilising the existing Kuhner dispersion cup. A thin slide valve was manufactured and fitted to the top face of the dispersion cup (Figure 7). The close fitting of the slider and fixed slides provided sealed chamber once the valve was closed. This design had the benefit of utilizing the original dispersion cup, nozzle and geometric locations of the dispersion and ignition points.

11

Figure 7: Slide valve dispersion chamber Test procedure: In common with the first design, the powder, dispersion chamber assembly and balance were placed in the glovebox. The air isolation valve was closed and the slide valve opened. Nanopowder was then weighed, placed in the dispersion cup and the powder was safely sealed within the cup by closing the slide valve. The assembly was removed from the glovebox, transferred and fitted to the MIKE3 apparatus and the compressed air line was connected to the lower chamber. The test procedure normally used in the operation of the Kuhner test apparatus was then followed in accordance with the general principles of BS EN 13821 (2002). Ignition energy levels used were 1 mJ, 3 mJ, 10 mJ, 30 mJ, 100 mJ, 300 mJ and 1000 mJ.

12

5 5.1

FIRE HAZARDS

FIRE HAZARDS IN NANOPOWDERS

With highly mobile, easily dispersed materials it is difficult to draw boundaries between ‘fire’ hazards, ‘deflagration’ hazards and ‘explosion’ hazards. ‘Fire’ is normally understood to be the ignition and combustion of material in bulk - as layers or within packages. Limiting consideration to these circumstances, there are three major fire issues that might arise with nanopowders: •

New ignition routes due to the presence of bulk nanopowders – e.g. the possibility of spontaneous combustion.

•

The possibility of enhanced burning rates in bulk nanopowders compared with solids or larger particulates and powders.

•

The potential for fire-induced air flows to lift bulk materials into a dispersed cloud while providing an ignition source for a nanopowder explosion.

The particular hazards in any fire situation are also threefold: •

heat, toxic gases and smoke that might lead to injury or death,

•

heat and smoke that might cause property damage

•

growth from a small, initial fire to a larger, more hazardous one.

At present, nanopowders other than carbon black are high value materials produced in relatively small quantities. Even ‘large’ amounts in storage are limited to kilogram quantities or less. In terms of fire load, the amount of combustible material present, the packaging of many nanopowders (plastics, cardboard, shrinkwrap etc) and the storage infrastructure (pallets, equipment, cables, even paints) are likely to represent the main hazard to life and to property. If nanomaterial combustion is relatively slow, such small quantities will not release energy quickly enough to be an unusual fire threat. On the other hand, nanopowders may exhibit a much more rapid combustion than equivalent materials at larger particle sizes. This means that they might release energy at a much faster rate, tending asymptotically to the explosion scenario. It seems unlikely that the ‘fire’ end of this range would cause significant harm with small amounts of nanopowders. 5.2

PRACTICAL ASSESSMENT OF FIRE PROPERTIES

Current nanopowders are typically highly mobile, requiring continuous containment to prevent losses even through diffusion into the air. This high mobility means that nanomaterials within process equipment would not be able to build up on surfaces without undergoing agglomeration or other physico-chemical changes, which would – in turn - affect their fire properties. Unchanged nanopowders are unlikely to “settle out” even in still conditions, and much less so in 13

any moving gas or air flows. Should nanomaterials escape from a process into a more open building (factory/storage) environment, the normal air movement present in large structures will help to reinforce their natural mobility and it is even less likely that any significant amount of nanomaterials could form a combustible mass or layer. The current high intrinsic value of almost all nanopowders means that there are also financial drivers on producers and handlers to minimise any losses – again reducing the likelihood of a significant amount of material being allowed to build up in process or storage. To pose a fire risk, nanopowders need to be present in layers containing many grams, even kilograms, of material. For the present, at least, such quantities will only be present for materials in storage. Any fire assessment therefore needs to address nanomaterials in storage conditions. The important fire properties (rate of heat production, rate of fire spread and formation of toxic combustion products) all include processes that are scale dependent. For example, the effect of thermal radiation is important in the spread of fire but is dependent on the square of the distance. This means that small-scale testing does not reproduce the effects of a full-scale fire. Consequently, fire tests need to be carried out at a representative scale. Nanopowders that are likely to be important in the future are currently only available in limited quantities at relatively high cost. The health and environmental risks associated with nanomaterials have not been completely defined; despite considerable ongoing effort in the area of nanotoxicology, many unknowns remain and it is recommended that a precautionary approach is taken when handling nanomaterials in research and manufacturing workplaces (Seaton and Tran, 2009). The combination of high cost and potential risk means that only the minimum necessary amount of work should be undertaken with nanopowders. Given the current low risk of harm from nanomaterial fires simply as a result of the low quantities present in any given location, it is difficult to justify fire testing to further quantify that risk.

14

6 6.1

ELECTROSTATIC PROPERTIES

ELECTROSTATIC CHARGE

The likelihood of accumulation of electrostatic charges and voltages in an industrial process increases with increasing electrical resistivity of powder (Eckhoff, 1997). With regard to dusts that penetrate into electrical and electronic equipment, the chance that they give rise to short circuits and equipment failure, increases as the dust resistivity decreases. In both situations the generation of ignition sources may result. BS EN 61241-2-2:1996 states that a dust with electrical resistivity equal to or less than 103 Ωm is considered to be a conductive dust and a dust with electrical resistivity greater than 103 Ωm is considered to be a non-conductive dust. The CCPS (2005) publication reports that resistivities above 108 Ωm are indicative of the potential for significant electrostatic charging. Increased humidity in the atmosphere where powders are handled may help to reduce the potential for electrostatic discharges by increasing the conductivity of the powders. PD CLC/TR 50404 (2003) divides powders into 3 groups depending on their volume resistivity: a) low resistivity powders with volume resistivities up to about 106 Ωm, b) medium resistivity powders in the range 106 Ωm to 1010 Ωm and c) high resistivity powders with resistivities of 1010 Ωm and above. NFPA77 (2007) has a similar 3category classification for powder resistivity. They indicate that low resistivity powders with resisitivities of up to 108 Ωm can become charged during flow. Powder resistivities within the range 108 to 1010 Ωm are considered to be medium resistivity powders. High resistivity powders have resistivities greater than 1010 Ω m. The significance of resistivity is discussed in NFPA 77, the main points being: a) Low resistivity powder can become charged during flow but rapidly dissipates when conveyed into a conductive grounded vessel. Powder conveyed into non-conductive container can result in an incendive spark. b) For medium resistivity powders coming to rest in bulk, the retained charge depends on the resistance between the powder and ground. Medium resistivity powders do not produce bulking brush discharges or sparks. c) High resistivity powders do not produce spark discharges themselves. However, they can produce corona, brush, bulking brush, and propagating brush discharges. Electrostatic charges are generated on solids during transport due to the rubbing of particles against particles or particles against equipment and the internal surfaces of pipes (CCPS, 2005). The charging of particles may take place independently of whether they are conductive or insulating materials. Equipment may also become charged by virtue of coming into contact with charged particles. However, an electrostatic charge by itself does not necessarily represent an ignition hazard – this only exits when the charge is sufficiently high that a discharge occurs due to the breakdown of a high voltage electric field. The criteria described above relates to micron-scale powders. These parameters have been used to assess the potential for charge build-up in a range of nanopwders, and the chance of nanopowders becoming potential ignition sources. Two test rigs have been constructed to assess charging and resistivity.

15

6.2

CHARGING TEST

In order to measure and demonstrate electrostatic charging of nanopowders in a realistic situation, a test apparatus has been constructed capable of generating particle movement and measuring the charge build-up. This incorporates a 25 mm diameter closed-loop continuous circulating PVC pipe incorporating a fan system to pneumatically circulate the powder. The overall length of the loop is approximately 1000 mm. A charge measuring station is incorporated into the pipe. Nanopowder was introduced through a valved port to be circulated within the loop. The aim is for any charge developed to be measured at an intermediate point in the loop. The system was developed, initially using four in-line fans, but later replaced with a centrifugal fan driven by a 55W single-phase motor (Figure 8).

Figure 8: charge test apparatus A number of techniques were employed to measure the deposition of charge at the measurement station during the commissioning of the test equipment. The technique that gave the most reliable results was to connect a capacitor between the measurement station and Earth; the voltage developed across the measurement capacitor is related to the charge accumulated on it. A number of measurement capacitor values were used during the tests, in each case the capacitance was accurately measured using a Wayne Kerr model 6425 LCR multi bridge. The voltage across the measurement capacitor was measured using a Keithley model 6514 electrometer. The data was captured using a PC running National Instruments’ Labview SignalExpress software and a National Instruments model USB-6221 data logger. The data logger captured data from the analogue output of the electrometer at a rate of 500Hz. Due to the high impedance of the electrometer it was necessary to provide an Earthed screen of wire mesh around the charging loop to control electrical noise. The relationship between the voltage across the capacitor and the supplied current is given by:

I =C

dV dt

Where I is the charging current (amps), dV/dt is the rate of change of voltage on the capacitor and C is the capacitance of the measurement capacitor (farad). 16

The rate of change of voltage in the powder charging cycle can be determined from the graph of voltage against time. Hence the streaming current can be calculated. Each powder was tested using a new, clean set of pipes. Before the test equipment was loaded with powder the resistance between the measurement station and earthed casing of the fan was measured using a Robin model 3075DL insulation and continuity meter. In each case the resistance was found to be greater than 2000 MΩ at 500V. The fan speed was set at the start of the tests and was not changed. The velocity of the air within the pipe was 28 m/s (measured using a TSI Airflow instrument) and is assumed to have been constant throughout the testing. The ambient environmental conditions were measured using a Rotronics Hygropalm 0 portable temperature and humidity meter. 6.3

RESISTIVITY TEST

Equipment has been designed and manufactured specifically to measure the bulk resisitivity of nanopowders in layers. The design is based on the equipment described in BS EN 61241-22:1996, but has been designed on a smaller scale to allow testing of small quantities of nanopowder. The equipment consists of a PTFE test cell (Figure 9) with two stainless steel bars. The bars have nominal cross sectional dimensions of 25 mm x 25 mm and are 25 mm long. These formed the electrodes and are located with a 10 mm long gap in a PTFE block. The block is 75 mm long x 90 mm wide with a channel in which the electrodes were located. Each electrode has a terminal connection for a calibrated resistance meter. Operator contact with nanopowders during these tests was minimised by locating the test cell in sealed test chamber within a glove box. Where necessary the atmosphere within the glovebox was inerted using argon. The humidity of the test chamber was adjusted using a desiccant and was monitored using a humidity meter. A weighed quantity of nanopowder was deposited within the cell and excess powder was removed by running a straight edge along the top of stainless steel electrodes. The bulk density of the sample was calculated by weighing the deposited sample. The electrodes were connected to the resistance meter and the resistance was measured over a range of humidity levels. The resistivity was calculated in accordance with the principles described in BS EN 61241-22:1996 and is described as follows. Where Ro is greater than 10 Rs, and the resistivity of the dust is calculated from the equation: ρ= 0.001 Rs [H × W/L] where ρ is the resistivity in Ω m Ro is the resistance of empty test cell in Ω Rs is the resistance of the filled test cell in Ω; H is the height of the electrode in mm; W is the length of the electrode in mm; L is the space between electrodes in mm. 17

A range of nanopowders was assessed and, for comparison purposes, a number of micron-scale powders were also tested.

Figure 9: Resistivity test cell

18

7

TEST MATERIALS

Nanopowder test materials were obtained including metal powders and organic materials and are listed in Table 2. Micron-scale powders used in the tests are listed in Table 3. Initial tests were done using Lycopodium, a well-established micron-sized standard test powder. This was supplied by Fluka Reference 19108 (Explosion Safety Unit Sample No EC/026/08. Intrinsiq Materials Ltd (previously Qinetiq Nanomaterials Ltd) supplied a quantity of aluminium nanopowder from their early production runs in which some carbon contamination had occurred. The material was useful for early commissioning of the test apparatus. Intrinsiq Materials Ltd also supplied commercial grade aluminium nanopowder. This material is passivated during production by exposing the material to an oxygen/argon atmosphere. This produces a partially oxidised material such that it can be exposed to the atmosphere during experiments without the risk of rapid oxidation and burning.

19

Table 2: Nanopowders Material

HSL Sample Number

Iron (passivated)

EC/147/07 Nanostructured & Amorphous 0266JY Materials, Inc. 16840 Clay Road, Suite 113, Houston, TX 77084, USA

Copper (passivated)

Supplier

Supplier’s Reference

EC/148/07 Nanostructured & Amorphous 0296JY Materials, Inc. 16840 Clay Road, Suite 113, Houston, TX 77084, USA

Information from supplier

Iron powder (Fe) (partially passivated) Purity: 99.5% (metal basis), APS: 25 nm SSA: 40 - 60 m2/g Color: black Morphology: spherical Bulk density: 0.10 - 0.25 g/cm3 True density: 7.87 g/cm3 Copper (Cu) (partially passivated) Purity: 99.8% (metal basis) APS: 25 nm SSA: 30 - 50 m2/g Color: black brown Morphology: spherical

Silicon Carbide

EC/159/07 Nanostructured & Amorphous 4620KE Materials, Inc. 16840 Clay Road, Suite 113, Houston, TX 77084, USA

Bulk density: 0.15 - 0.35 g/cm3 True density: 8.94 g/cm3 Purity 97.5%; APS 45-55 nm; colour Grey white: Bulk Density 0.068 g/cm3; True density 3.22g/cm3; Morphology spherical; Specific surface area 70-90 m2/g

Zinc Oxide

EC/149/07 Nanostructured & Amorphous 5810MR Materials, Inc. 16840 Clay Road, Suite 113, Houston, TX 77084, USA

Zinc Oxide (ZnO) Purity 99.5%; APS: 20nm; SSA: 50 m2/g; Colour: milky white; Morphology: nearly spherical; Bulk density: 0.3-0.45 g/cm3; True density: 5.606 g/cm3

Iron Oxide

EC/150/07 Nanostructured & Amorphous 2652FY Materials, Inc. 16840 Clay Road, Suite 113, Houston, TX 77084, USA

Iron Oxide (Fe3O4); Purity 99.5%; APS 15-20 nm; Colour: black; Morphology : spherical; Bulk density : 0.8-0.9 g/cm3; True density 4.8-5.1 g/cm3

Zirconium Oxide

EC/151/07 Nanostructured & Amorphous 5930LQ Materials, Inc. 16840 Clay Road, Suite 113, Houston, TX 77084, USA

Zirconium Oxide (ZrO2); Purity 99.95%; APS: 29-68 nm; SSA: 15-35 m2/g; Colour: white; Morphology: spherical; Bulk density: 0.74 g/cm3; True density 5.89 g/cm3

Zinc

EC/152/07 Nanostructured & Amorphous 0303WF Materials, Inc. 16840 Clay Road, Suite 113, Houston, TX 77084, USA

Zinc (Zn); Purity 99.5%; APS : 130nm; SSA: 6.4 m2/g; Colour : Grey; Morphology: spherical; Bulk density: 0.70.85 g/cm3; True density 7.14 g/cm3

Multi-walled nanotubes

Carbon nanofibres

carbon EC/153/07 Nanostructured & Amorphous 1229YJ Materials, Inc. 16840 Clay Road, Suite 113, Houston, TX 77084, USA

MWNT Purity: > 95% ; Outside diameter: 20-30 nm; Inside diameter: 5-10 nm; Length: 10-30 um; SSA: ~ 110 m2/g; Colour: black; True density: ~2.1 g/cm3

EC/158/07 Nanostructured & Amorphous 1190JN Materials, Inc. 16840 Clay Road, Suite 113, Houston, TX 77084, USA

Carbon nanofibers; Purity 95%; Outside diameter: 80-200 nm; Core diameter: 1-10 nm; Length 0.5-20 um;BET: 2535 m2/g; Electronic Resistivity: 0.75-0.1 Omu.cm (20180MPa); Colour: black; Bulk density: 0.06-0.08 g/cm3; True density: 1.9 g/cm3

20

Table 2 (continued): Nanopowders Carbon Nanofibre

EC/042/08

Pyrograph Products Inc

PR-19-SLD

Carbon Nanofibre

EC/116/08

Applied Sciences Inc.

Carbon Nanofibre

EC/117/08

Applied Sciences Inc

Aluminium

EC/104/08

Intrinsiq Materials Ltd

PR-25-XTPS-AM PR-24-XTHHT-AM QNA 2607

Aluminium

EC/011/09


QNA 2798

Aluminium

EC/060/07


Carbon Black

EC/076/07

Cabot

Ref: PQN301 Batch 17 Monarch 800

Material

Supplier

Supplier’s Reference

Aluminium

HSL Sample Number EC/085/08

Alpoco

Iron

EC/003/09

Ronald Britton Ltd

Zinc oxide Copper Zirconium oxide Carbon Lycopodium Zinc Stearate Coal Toner

EC/112/08 EC/004/09 EC/111/08 EC/113/08 EC/026/08 EC/118/08 EC/120/08 EC/122/08

Fluka Ronald Briton Ltd Acros Organics Sigma Aldrich Fluka

Batch 137X Code B10121 Batch 831302 96479

Diameter 100-200 nm Length 30-100 micron Diameter 70-200 nm Length 2-5 micron Diameter 70-200 nm Length 2-10 micron Nominally 100nm Minimum 60% Al, remainder Aluminium oxide surface layer. Trace elements Fe, Na, Zn, Ga. Nominally 210 nm Minimum 60% Al, remainder Aluminium oxide surface layer. Trace elements Fe, Na, Zn, Ga. Estimated 73-109 nm 17 nm immediately after formation in the combustion process.

Table 3: Micron-scale powders Information from supplier 8-

Specification 99.7% Al, 8 micron. 100% 1000 mJ

EC/116/08

Carbon nanotubes

> 1000 mJ

EC/147/07

Iron nanopowder

< 1 mJ

EC/152/07

Zinc nanopowder

3 – 10 mJ

EC/148/07

Copper nanopowder

> 1000 mJ

35

Figure 26: Iron nanopowder

Figure 27: Zinc nanopowder 36

8.5

ELECTROSTATICS

8.5.1

Resistivity test results

8.5.1.1

Aluminium

Test data for aluminium nanopowder and micron-scale aluminium powder is presented in Table 11. and in Figure 28.

Table 11: Aluminium resistivity Aluminium nanopowder EC/104/08 Relative Resistivity (Ω.m) humidity (%) Resistance (Ω)

Aluminium standard powder EC/085/08 Relative Resistivity (Ω.m) humidity (%) Resistance (Ω)

70.0

5.50E+07

3.44E+06

75.0

3.03E+03

1.89E+02

60.0

9.48E+07

5.93E+06

65.0

2.89E+03

1.81E+02

50.0

2.21E+08

1.38E+07

55.0

3.36E+03

2.10E+02

40.0

7.01E+08

4.38E+07

45.0

3.97E+03

2.48E+02

30.0

5.65E+10

3.53E+09

35.0

3.93E+03

2.46E+02

20.0

3.86E+11

2.41E+10

27.0

3.71E+03

2.32E+02

8.1

5.82E+12

3.64E+11

23.0

2.90E+03

1.81E+02

Bulk density 0.221 g/cm3

9.1

2.94E+03

1.84E+02

Bulk density 0.968 g/cm3

R e si st i v i t y v R e l a t i v e H u m i di t y f o r A l u m i ni um P owd e r s

Aluminium nanopow der EC/104/08

Aluminium standard pow der EC/085/08

1.00E+12 1.00E+11 1.00E+10 1.00E+09

Ω

1.00E+08 1.00E+07 1.00E+06 1.00E+05 1.00E+04 1.00E+03 1.00E+02 1.00E+01 1.00E+00 0.0

10.0

20.0

30.0

40.0

50.0

60.0

70.0

80.0

R e l a t i v e H umi di t y ( %)

Figure 28: Aluminium resistivity

37

90.0

100.0

8.5.1.2

Iron

Test data for iron oxide nanopowder, iron nanopowder and micron-scale iron powder is presented in Table 12. and in Figure 29. Table 12: Iron resistivity Iron oxide nanopowder EC/150/07 Relative Resistance (Ω) Resistivity (Ω.m) humidity (%) 75.0 3.85E+09 2.41E+08 65.0 3.87E+09 2.42E+08 55.0 3.96E+09 2.48E+08 45.0 4.17E+09 2.61E+08 35.0 4.46E+09 2.79E+08 25.0 5.44E+09 3.40E+08 16.6 7.63E+09 4.77E+08 7.7 5.51E+10 3.44E+09 Bulk density 1.108 g/cm3

Iron standard powder EC/003/09 Relative Resistivity (Ω.m) humidity (%) Resistance (Ω) 65.0 51.5 3.22E+00 60.0 51.7 3.23E+00 55.0 51.8 3.24E+00 45.0 51.7 3.23E+00 35.0 51.8 3.24E+00 25.0 51.6 3.23E+00 5.0 51.2 3.20E+00 Bulk density 3.497 g/cm3

Iron nanopowder EC/147/07D (inert atmosphere) Relative Resistivity (Ω.m) humidity (%) Resistance (Ω) 7.8 1.45E+04 9.06E+02 Bulk density 0.606 g/cm3

Resistivity v Relative Hum idity for Iron Pow ders Iron nanopow der EC/147/07D Iron oxide nanopow der EC/150/07

Iron standard pow der EC/003/09

1.00E+10 1.00E+09 1.00E+08

Ω

1.00E+07 1.00E+06 1.00E+05 1.00E+04 1.00E+03 1.00E+02 1.00E+01 1.00E+00 0.0

10.0

20.0

30.0

40.0

50.0

60.0

R elat ive Humid i t y ( %)

Figure 29: Iron resistivity

38

70.0

80.0

90.0

100.0

8.5.1.3

Zinc

Test data for zinc oxide nanopowder, zinc nanopowder and micron-scale zinc powder is presented in Table 13 and in Figure 30. Table 13: Zinc resistivity Zinc oxide nanopowder EC/149/07

Zinc oxide standard powder EC/112/08 Relative humidity (%) Resistance (Ω) Resistivity (Ω.m) 90.0 1.93E+09 1.21E+08 85.0 1.77E+09 1.11E+08 78.1 1.93E+09 1.21E+08 70.0 2.13E+09 1.33E+08 63.2 2.44E+09 1.53E+08 52.3 3.52E+09 2.20E+08 44.9 5.42E+09 3.39E+08 40.2 7.61E+09 4.76E+08 28.8 2.48E+10 1.55E+09 23.5 6.10E+10 3.81E+09 19.9 1.14E+11 7.13E+09 16.6 2.29E+11 1.43E+10 14.6 4.68E+11 2.93E+10 8.5 1.82E+13 1.14E+12 Bulk density 0.690 g/cm3

Relative Resistivity (Ω.m) humidity (%) Resistance (Ω) 76.7 2.23E+06 1.39E+05 66.0 3.00E+06 1.88E+05 60.0 3.18E+06 1.99E+05 54.0 3.71E+06 2.32E+05 50.0 3.93E+06 2.46E+05 43.5 4.30E+06 2.69E+05 40.0 4.87E+06 3.04E+05 35.7 5.24E+06 3.28E+05 25.7 7.68E+06 4.80E+05 22.2 1.11E+07 6.95E+05 18.9 1.72E+07 1.07E+06 17.4 2.52E+07 1.58E+06 15.6 3.50E+07 2.19E+06 14.9 5.01E+07 3.13E+06 8.7 9.83E+10 6.14E+09 Bulk density 0.523 g/cm3 Zinc nanopowder EC/152/07 (inert atmosphere) Relative Resistivity (Ω.m) humidity (%) Resistance (Ω) 5.7 6.00E+04 3.75E+03 Bulk density 1.220 g/cm3

Resistivity v Relative Humidity for Zinc Oxide powders Zinc oxide nanopowder EC/149/07

Zinc oxide standard powder EC/112/08

1.00E+13 1.00E+12 1.00E+11 1.00E+10

Resistivity (Ω.m)

1.00E+09 1.00E+08 1.00E+07 1.00E+06 1.00E+05 1.00E+04 1.00E+03 1.00E+02 1.00E+01 1.00E+00 0.0

10.0

20.0

30.0

40.0

50.0

60.0

Relative Humidity (%)

Figure 30: Zinc resistivity

39

70.0

80.0

90.0

100.0

8.5.1.4

Copper

Test data for copper nanopowder and micron-scale copper powder is presented in Table 14 and in Figure 31. Table 14: copper resistivity Copper nanopowder EC/148/07A (inert atmosphere) Relative humidity (%)

Resistance (Ω) 6.6

Resistivity (Ω.m)

3.75E+03

2.34E+02 Bulk density 0.523 g/cm3

Copper standard powder EC/004/09 Relative humidity (%)

Resistance (Ω)

Resistivity (Ω.m)

70.0

121.9

7.62E+00

60.0

120.7

7.54E+00

50.0

119.8

7.49E+00

45.0

116.3

7.27E+00

40.0

115.9

7.24E+00

30.0

115.4

7.21E+00

25.0

114.7

7.17E+00

10.0

113.6


Resistivity of Copper Pow ders versus Hum idity Copper nanopow der EC/148/07A

Copper standard pow der EC/004/09

1.00E+05

Ω

1.00E+04

1.00E+03

1.00E+02

1.00E+01

1.00E+00 0.0

10.0

20.0

30.0

40.0

50.0

60.0

R e l a t i v e H u m i di t y ( %)

Figure 31: Copper resistivity 40

70.0

80.0

90.0

100.0

8.5.1.5

Zirconium oxide

Test data for zirconium oxide nanopowder and micron-scale zirconium powder is presented in Table 15 and in Figure 32. Table 15: zirconium oxide Zirconium oxide nanopowder EC/151/07D Relative humidity (%)

Resistance (Ω) 70.0 60.0 50.0 40.0 30.0 23.0 8.8

Resistivity (Ω.m) 2.35E+07 2.58E+07 2.80E+07 3.34E+07 5.29E+07 1.64E+08 1.20E+09

1.47E+06 1.61E+06 1.75E+06 2.09E+06 3.31E+06 1.02E+07 7.51E+07 Bulk density 0.953 g/cm3

Zirconium oxide standard powder EC/111/08 Relative humidity (%)

Resistance (Ω) 75.0 70.0 65.0 60.0 50.0 40.0 28.3 22.5 18.3 16.9 7.5

Resistivity (Ω.m) 1.85E+07 1.88E+07 1.93E+07 1.96E+07 2.05E+07 2.21E+07 2.64E+07 3.98E+07 5.08E+07 5.92E+07 3.95E+10

1.16E+06 1.17E+06 1.20E+06 1.23E+06 1.28E+06 1.38E+06 1.65E+06 2.49E+06 3.18E+06 3.70E+06 2.47E+09 Bulk density 1.716 g/m3

Resistivity v Relative Humidity For Zirconium Oxide Powders Zirconium oxide nanopowder EC/151/07D

Zirconium oxide standard powder EC/111/08

1.00E+10 1.00E+09 1.00E+08

Resistivity (Ω.m)

1.00E+07 1.00E+06 1.00E+05 1.00E+04 1.00E+03 1.00E+02 1.00E+01 1.00E+00 0.0

10.0

20.0

30.0

40.0

50.0

60.0

70.0

Relative Humidity (%)

Figure 32: Zirconium oxide resistivity

41

80.0

90.0

100.0

8.5.1.6

Carbon

Test data for carbon nanomaterials and micron-scale carbon powder is presented in Table 16 and in Figure 33. Table 16: Carbon resistivity Multi-walled carbon nanotubes EC/153/07C Relative humidity (%)

Resistance (Ω)

Resistivity (Ω.m)

75.0

1.71E+02

1.07E+01

65.0

1.64E+02

1.03E+01

55.0

1.70E+02

1.06E+01

45.0

1.69E+02

1.06E+01

35.0

1.78E+02

1.11E+01

25.0

4.64E+01

2.90E+00

15.0

4.46E+01

2.79E+00

7.6

4.44E+01


Carbon nanofibres EC/158/07A Relative humidity (%)

Resistance (Ω)

Resistivity (Ω.m)

75.0

8.44E+02

5.28E+01

64.0

7.98E+02

4.99E+01

55.0

7.66E+02

4.79E+01

45.0

7.52E+02

4.70E+01

35.0

7.41E+02

4.63E+01

25.0

7.40E+02

4.63E+01

17.1

7.20E+02

4.50E+01

7.8

3.05E+01


Pyrograf III nanofibres EC/042/08 Relative humidity (%)

Resistance (Ω)

Resistivity (Ω.m)

75.0

8.36E+04

5.23E+03

65.0

8.39E+04

5.24E+03

55.0

8.58E+04

5.36E+03

45.0

8.88E+04

5.55E+03

33.0

8.73E+04

5.46E+03

25.0

9.20E+04

5.75E+03

19.6

9.23E+04

5.77E+03

9.8

2.48E+04


Carbon micron powder EC/113/08 Relative humidity (%)

Resistance (Ω)

Resistivity (Ω.m)

70.0

3.25

2.03E-01

60.0

3.19

1.99E-01

50.0

3.15

1.97E-01

40.0

3.11

1.94E-01

28.0

2.93

1.83E-01

21.5

2.8

1.75E-01

8.5

3.2

2.00E-01 Bulk density 0.577 g/cm3

42

Resistivity v Relative Hum idity for Carbon Pow ders Multi-w alled carbon nanotubes EC/153/07C Pyrograf III nanofibres EC/042/08

Carbon nanofibres EC/158/07A Carbon micron pow der EC/113/08

1.00E+04

Ω

1.00E+03

1.00E+02

1.00E+01

1.00E+00

1.00E-01 0.0

10.0

20.0

30.0

40.0

50.0

60.0

R el at i ve Humi d it y ( %)

Figure 33: Carbon resistivity

43

70.0

80.0

90.0

100.0

8.5.2

Charging test results

8.5.2.1

Commissioning

Base-line test As part of the commissioning tests on this equipment, a base-line test with no powder was initially carried out; this was done regularly during the tests to ensure the instrumentation was functioning correctly. Although this test was performed with new, clean pipes, surface and air contaminants seem to have contributed to the occurrence of some charging. An example of the plot of voltage against time is shown in Figure 34 and the test data is shown in Table 17. The airflow was maintained at a velocity of 28 m/s for the test programme. 100m 50m 0 -50m - 100m - 150m - 200m - 250m - 300m - 350m - 400m - 450m - 500m - 550m - 600m - 650m - 700m - 750m - 800m - 850m - 900m - 950m -1 -1.05 -1.1 -1.15 -1.2 -1.25 0

50

100

150

2 00

250

300

350

4 00

45 0 Time (s)

500

550

600

65 0

700

750

800

8 50

88 0.2

Figure 34 : Base-line test without powder

Table 17 : Base-line test without powder Rate of change of voltage (V/s) -6.29E-03

Capacitance (F) 1.02E-08

Ambient environmental conditions: Humidity: 67.0%RH Temperature: 19.9ºC

44

Current (A) -6.40E-11

Lycopodium (EC/026/08) To test the system, micron-scale lycopodium powder was tested. This was free flowing and did not aggregate or accumulate in any significant quantities. During testing a very fine layer of powder coated the inside wall of the pipes. Figure 35 shows the plot of voltage against time and Table 18 presents the test results. An initial period of high rate of change of voltage was noted in the first few seconds following the fan being started, the cause of this was not investigated. This was followed by a period of steady gradient. 11.5 11.25 11 10.75 10.5 10.25 10 9.75 9.5 9.25 9 8.75 8.5 8.25 8 7.75 7.5 7.25 7 6.75 6.5 6.25 6 5.75 5.5 5.25 5 4.75 4.5 4.25 4 3.75 3.5 3.25 3 2.75 2.5 2.25 2 1.75 1.5 1.25 1 750m 500m 250m 0 -250m 0

10

20

30

40

50

60

70

80

90

100

110

120

130 Time (s )

14 0

150

160

170

18 0

190

200

210

220

230

240

250 255.7

Figure 35: Lycopodium (EC/026/08) Table 18: Lycopodium (EC/026/08) Rate of change of voltage (V/s) 1.76E-01


45

Current (A) 6.77E-10

8.5.2.2

Charging tests

Micron-scale zinc oxide (EC/112/08) The micron scale zinc oxide powder was observed to be highly cohesive; small quantities could be compressed together to form a layer on the surface of a spatula. A mass of 0.1g of powder was tested. Charge was generated for less than 1.5 seconds. The powder quickly accumulated around any obstructions and rapidly removed the powder out of circulation; this was particularly apparent where the plastic pipe was connected onto a metal tube at the powder injection port. Some powder was also deposited on the inner wall of the pipes. The test results are shown in Table 19.

Table 19: Micron-scale Zinc Oxide (EC/112/08) Rate of change of voltage (V/s) 8.12E-01



46


Zinc oxide nanopowder (EC/149/07) Zinc oxide nanopowder had a very fine consistency with very little cohesion under compression. A mass of 6g of powder was tested. Charging was observed for the whole duration of the test (Figure 36). A number of phases of charging each with different gradients are apparent from the data. There was little powder accumulation observed but a significant layer of powder formed at the inner walls of the pipe. A good proportion of the powder was observed resting in the bottom of the pipes following the test. The test results are shown in Table 20. 8 60m 8 40m 8 20m 8 00m 7 80m 7 60m 7 40m 7 20m 7 00m 6 80m 6 60m 6 40m 6 20m 6 00m 5 80m 5 60m 5 40m 5 20m 5 00m 4 80m 4 60m 4 40m 4 20m 4 00m 3 80m 3 60m 3 40m 3 20m 3 00m 2 80m 2 60m 2 40m 2 20m 2 00m 1 80m 1 60m 1 40m 1 20m 1 00m 80m 60m 40m 20m 0 -20m -40m 0

20

40

60

80

1 00

120

14 0

160

18 0

200

220 Time (s)

2 40

260

2 80

300

32 0

340

360

380

400

4 20

43 3.1

Figure 36: Zinc oxide nanopowder (EC/149/07)

Table 20: Zinc oxide nanopowder (EC/149/07) Rate of change of voltage (V/s) 8.92E-03



47


8.5.2.3

Zirconium Oxide

Micron-scale zirconium oxide (EC/111/08) The powder had a very fine consistency with very little cohesion under compression. A mass of 1g of powder was tested. A rate of charging greater than anticipated was observed; the voltage across the capacitor rapidly increased to a value greater than could be measured. The powder was deposited in a thick layer on the inner wall of the pipe. Following the test the resistance between measuring station and the Earthed fan casing was measured as greater than 2000MΩ. The test results are shown in Table 21. Table 21: Micron-scale zirconium oxide (EC/111/08) Rate of change of voltage (V/s) 4.95E+00




Zirconium oxide nanopowder (EC/151/07) The powder had a very fine consistency with very little cohesion under compression. A quantity of 1g of powder was tested. The powder accumulated at the pipe joints forming a significant build-up. Charging was recorded for approximately 2 seconds. There was little powder was deposited on the inner wall of the pipe. Further quantities of powder were added to try to overcome the problems of powder build and hence reduced powder circulation. A maximum of 6g of powder were used, these tests resulted in very similar behaviour and charging magnitudes. The test results are shown in Table 22.

Table 22: Zirconium oxide nanopowder (EC/151/07) Rate of change of voltage (V/s) -1.84E-01



48


8.5.2.4

Carbon

Micron-scale carbon (EC/113/08) The powder was very cohesive; this was evident in the difficulty experienced in releasing the powder from the powder loading port. A quantity of 3g of powder was tested. The voltage across the capacitor rapidly increased to a value greater than could be measured. The powder was deposited in a significant layer on the inner wall of the pipe. A plot of the test is shown in Figure 37 and the test results are shown in Table 23. It was noted that the voltage on the capacitor returned to 0V within 1 second once the fan was stopped. The resistance between measuring station and the Earthed fan casing was measured as 0.50MΩ. The carbon deposits on the inner walls of the pipe has significantly reduced the resistance to Earth.

10 .74 1 0.5 10 .25 10 9 .75 9.5 9 .25 9 8 .75 8.5 8 .25 8 7 .75 7.5 7 .25 7 6 .75 6.5 6 .25 6 5 .75 5.5 5 .25 5 4 .75 4.5 4 .25 4 3 .75 3.5 3 .25 3 2 .75 2.5 2 .25 2 1 .75 1.5 1 .25 1 75 0m 50 0m 25 0m 0 -263.4m 1 1:36:55.739

11:36:57.739

11:36 :59.739

11:37:01.739

11:37:03. 739 Time (s)

11:37:05.739

11:37:07.739

11:37:09.739

11:37:11.544

Figure 37: Micron Scale Carbon (EC/113/08)

Table 23: Micron Scale Carbon (EC/113/08) Rate of change of voltage (V/s) 1.28E+01



49


Carbon Pyrograf III nanofibres (EC/042/08) The powder was observed as being made up of mixed sized particles from fine dust to agglomerated particles of approximately 2mm in size. Powder could not be aggregated further when compressed. A quantity of 2g of powder was tested. A plot of the test is shown in Figure 38 and the test data is shown in Table 24. A continuous flow of powder was observed, there was no apparent powder accumulation. A significant layer of powder was deposited on the inner wall of the pipe. A quantity of mainly larger sized particles where present on the bottom of the pipe once the fan was stopped. 100m 50m 0 -50m -100m -150m -200m -250m -300m -350m -400m -450m -500m -550m -600m -650m -700m -750m -800m -850m -900m -950m -1 -1.05 -1. 1 -1.15 -1. 2 -1.25 -1. 3 -1.35 -1. 4 -1.45 -1. 5 -1.55 14: 00:36.256

14:01:26 .256

14:02:16.256

14:03:06.256

14:03:56.2 56

14:04:46.256

14:05: 36.256

14:06:26.25 6

14 :07:16.256

Time (s)

Figure 38: Carbon Pyrograf III nanofibres (EC/042/08) Table 24: Carbon Pyrograf III nanofibres (EC/042/08) Rate of change of voltage (V/s) -3.42E-02



50


14:08 :11.156

Carbon nanofibres (EC/158/07) In common with sample EC/042/08, the powder was observed as being made up of mixed sized particles from fine dust to agglomerated particles of approximately 2mm in size. Powder could not be aggregated further when compressed. A quantity of 1g of powder was tested. During the test, a continuous flow of powder was observed but a layer of powder was deposited on the inner wall of the pipe. A quantity of mainly larger sized particles where present on the bottom of the pipe once the fan was stopped. The plot of the test is shown in Figure 39 and the test data is presented in Table 25. It was observed that the voltage across the capacitor dropped away rapid during the test. The resistance between measuring station and the Earthed fan casing was measured as 0.12MΩ. The carbon deposits on the inner walls of the pipe has significantly reduced the resistance to Earth. 20m 0 -20m -40m -60m -80m -100m -120m -140m -160m -180m -200m -220m -240m -260m -280m -300m -320m -340m -360m -380m -400m -420m -440m -460m -480m -500m -520m -540m -560m -580m -600m -620m -640m -660m -680m -700m -720m -740m -760m -780m -800m -820m 0

10

20

30

40

50

60

70

80

90

100

110

120

1 30 140 Time (s )

150

160

170

180

19 0

200

210

220

230

240

25 0

2 60

26 8.7

Figure 39: Carbon nanofibres (EC/158/07)

Table 25: Carbon nanofibres (EC/158/07) Rate of change of voltage (V/s) -1.16E-02



51


Carbon multi-walled nanotubes (EC/153/07) The powder was observed as being made up of mixed sized particles from fine material of approximately 2mm in diameter. Powder could not be aggregated further when compressed. A quantity of 1g of powder was tested. A continuous flow of powder was observed. A significant layer of powder was deposited on the inner wall of the pipe and a quantity of mainly larger sized particles where present on the bottom of the pipe once the fan was stopped. During the test there was audible evidence of electrostatic discharges. However, there was no apparent visible evidence of discharges. Following the test the resistance between measuring station and the Earthed fan casing was measured as 0.15MΩ. A plot of the test is shown in Figure 40 and the test data is presented in Table 26. 20m 0 -20m -40m -60m -80m -100m -120m -140m -160m -180m -200m -220m -240m -260m -280m -300m -320m -340m -360m -380m -400m -420m -440m -460m -480m -500m -520m -540m -560m -580m -600m -620m -640m -660m 16:23:15.287

16 :23:35.287

1 6:23:55.287

16:24:15.287

16:24:35.287

16:24:55.28 7

16:25:15.2 87

16:25:35. 287 Time (s)

16:25:55 .287

16:26:1 5.287

16:26 :35.287

16:2 6:55.287

16:27:15.287

16:27 :48.987

Figure 40: Carbon multi-walled nanotubes (EC/153/07)

Table 26: Carbon multi-walled nanotubes (EC/153/07) Rate of change of voltage (V/s) -2.51E-02



52


8.5.2.5

Copper

Micron-scale copper (EC/004/09) The powder had a very fine consistency with very little cohesion under compression. A quantity of 5g of powder was tested. A continuous flow of powder was observed; there was no apparent powder accumulation. A very fine layer of powder was deposited on the inner wall of the pipe. The test data is presented in Table 27.

Table 27: Micron-scale copper (EC/004/09) Rate of change of voltage (V/s) 5.65E-02




Copper nanopowder (EC/148/07) The powder was naturally aggregated into approximately 0.5mm material. The powder could not be aggregated further when compressed. A quantity of 2.5g of powder was tested. The powder was initially observed to flow reasonably well. The inner wall of the pipe was coated with a fine layer of powder shortly after starting the test. When the test was complete there was little evidence of any free powder remaining in the pipes. Following the test the resistance between measuring station and the Earthed fan casing was measured as greater than 2000MΩ. The test data is presented in Table 28.

Table 28: Copper nanopowder (EC/148/07) Rate of change of voltage (V/s) -4.89E-02



53


9 9.1

DISCUSSION

EXPLOSION TEST EQUIPMENT

A 2 litre explosion test facility has been established. Equipment has been manufactured and where necessary existing equipment has been modified and configured to allow the safe handling and testing of nanopowders. Sealed systems have been designed for use in conjunction with a glovebox that have allowed nanopowders to be weighed and handled under an inert atmosphere. This has enabled non-oxidised materials to be safely handled and held under inert conditions until the point of ignition. The 2 litre explosion test apparatus has been used to measure the explosion pressure and rates of pressure rise of nanopowders using a set of conditions established and calibrated against a range of dusts tested in both the 2 litre and 20 litre vessels. The small volume of the 2 litre vessel has reduced the quantity of test material required for the test procedure. Typically, 50 – 100 g of material is required for a series of tests covering a wide range of dust concentrations. This is approximately 10% of the material requirements for a standard 20 litre sphere test. The small volume of the test chamber was a concern in that quenching of the flame at the vessel wall was possible and could lead to lower combustion rates. This proved to be the case and the rates of pressure rise were lower than might be expected from a small vessel. However, a scaling factor has been obtained from the test data and equivalent KSt values have been obtained. The maximum explosion pressures were generally comparable with those measured in the 20 litre sphere. The minimum ignition energy of nanopowders was measured using a modified Kuhner MIKE3 test apparatus. The equipment was modified to allow nanopowder to be safely handled within the test apparatus. The modifications included enclosing the dispersion cup with a slide valve to fully contain the nanopowder and allow handling within a glovebox. Argon was used for inerting the material. The full test chamber assembly was configured such that it was removable as a complete sealed assembly for cleaning purposes.

9.2

EXPLOSION SEVERITY AND MIE

A comparison of the nanopowder data generated in this test programme with published micronscale powder data (BIA-Report 1997) is summarised in Table 29. Generally, the KSt values of nanopowders are similar to conventional micron-scale powders. The Pmax values are a little lower due to flame quenching effects at the vessel wall. Surprisingly, the large surface to volume ratio has not produced greater reactivity than the equivalent material at micron-scale. However, the minimum ignition energies of some nanopowders have been lower than the equivalent material at micron-scale. This may be explained by consideration of dispersion and agglomeration (Preining, 1998). Although powders are initially well dispersed in the 2 litre chamber, the rate of re-agglomeration may be such that nanopowders are dispersed only for a very short duration and rapidly begin to re-agglomerate. Thus, combustion may well take place initially with dispersed particles but some degree of re-agglomeration of the particles may take place before combustion has been completed. This hypothesis would need to be checked by sampling the dispersed dust cloud and measuring particle size using equipment that measures the size distribution and number concentration of fast dynamic aerosols. Equipment needed for measuring the particle size of the dispersed dust cloud was considered during this project. The current bench-mark instrument for measurement of nanoparticles in 54

their airborne state is considered to be the Scanning Mobility Particle Sizer (SMPS). This instrument measures a size and number distribution from around 5 nanometres up to 1 micron in diameter. The SMPS, comprises a Condensation Particle Counter (CPC) and a Differential Mobility Analyser (DMA). The latter component invariably employs a radioactive source to produce a known charge distribution on the sampled aerosol through ionisation of the sampled air. However, the SMPS cannot be used to obtain a particle size distribution in anything like realtime. Full particle size data is derived from a series of scans during which a varying voltage is applied to the selection electrode in the Differential Mobility Analyser. In general, instruments of the SMPS type require around three minutes to complete a full particle size scan. This means that the information gathered by the SMPS will be spurious if the concentration and size range of the aerosol being sampled are dynamic and not constant. This is particularly disadvantageous when measuring nanoparticles dispersed into the air during explosion testing. Under these circumstances, the distribution of sizes will be extremely transitory due to the tendency of the primary particles to rapidly agglomerate. Therefore the SMPS equipment could not be used for assessment of the particle size and agglomeration. An instrument is required that will capture the transient particle size distribution. The instrument identified for this work is a Fast Mobility Particle Sizer Spectrometer. A sample in near real-time is taken and is therefore an ideal tool to measure an aerosol composed of nanoparticles, either in the laboratory or in the field, under rapidly changing conditions. It is recommended that further work is done to assess the particle size/agglomeration of the dust cloud at the point of ignition using this instrument. In common with micron-scale powders the maximum rate of pressure rise and maximum explosion pressure of nanopowders sometimes occurred at different dust concentrations. The metal dusts tend to have relatively high optimum dust concentrations compared with carbon nanopowders, for example the optimum dust concentration for aluminium is within the range 2000 – 3000 g/m3 whereas carbon is typically within the range 400 –1500 g/m3. 9.2.1

Aluminium

The aluminium nanopowders have demonstrated explosion properties comparable with the micron-scale powders. Published data (BIA, 1997) of micron-scale aluminium powders typically have KSt values within the range 300 – 700 bar m/s, and have maximum explosion pressures within the range 7 – 12 bar. KSt values of 449 bar m/s for 210 nm aluminium nanopowder and 536 bar m/s for 100 nm nanopowder powder are well within this range of values. As might be expected, the finer aluminium nanopowder had the higher KSt value. Other research workers (Nanosafe, 2008) have found the KSt of some aluminium nanpowders to decrease with decreasing particle size. This has been attributed to the thin oxide layer wrapping the passivated nanoparticles; this may make them less explosible than micron-scale aluminium nanopowders. The aluminium nanopowders are easily ignited, burning with a highly incandescent flame and are firmly in the St3 Dust Group. Aluminium nanopowders, as with micron-scale aluminium powders, have very severe explosion characteristics. The minimum ignition energies of aluminium nanopowders were comparable with micron-scale aluminium with MIE < 1 mJ. The aluminium nanopowders are therefore very sensitive to ignition by electrostatic discharge and extreme caution is required when handling these powders.

55

9.2.2

Zinc

Visually, zinc nanopowder burned in a similar manner to aluminium, producing a very bright flame. The maximum explosion pressure was 5.6 bar, the maximum rate of pressure rise was 377 bar/s resulting in an equivalent KSt value of 101 bar m/s. The nanopowder is thus classified as an St1 dust. The zinc nanopowder characteristics are typical of published micron-scale zinc powders (BIA 1997). The minimum ignition energy was 3 – 10 mJ and was considerably lower than published data for micron-scale zinc powder; 100-1000 mJ and 300-1000 mJ have been quoted. The ignitability of zinc nanopowder is therefore greater than conventional micron-scale zinc powder. 9.2.3

Iron

Iron nanopowder produced an explosion pressure of almost 3 bar and a rate of pressure rise of 68 bar m/s resulting in an equivalent KSt value of 18 bar m/s. This was lower than published data for micron-scale iron where explosion pressures 5.1 - 5.2 bar and KSt values of 41 - 50 bar m/s have been published. Iron nanopowder is clearly an St1 dust. The minimum ignition energy of iron nanopowder was less than 1 mJ. Although the explosion data is not as severe as micron-scale powder, the low MIE value is lower than micron-scale iron powder and demonstrates it is potentially a very ignitable nanopowder. 9.2.4

Copper

Copper nanopowder proved to have very low explosion characteristics to the extent that it is considered non-explosible. This is consistent with micron-scale copper powder; test data in the BIA database indicates micron-scale copper powder is non-explosible. 9.2.5

Carbon

The carbon nanopowders generally had a very agglomerated appearance and had poor handling characteristics in that they would not flow in the manner of free-flowing powders. KSt values were within the range 17-158 bar m/s and are therefore classified as St1 dusts and are weak to moderately explosible. The difference in the reactivity of the two extremes may be due to the length of the fibres. Sample EC/042/08 was the lowest rated carbon nanofibre and had relatively long fibres; 30-100 micron long with a diameter of 100-200 nm. The highest rated material was carbon nanofibre sample EC/116/08; this had a diameter of 70 – 200 nm but a significantly shorter length of 2-5 micron. Therefore, as might be expected, the shorter carbon nanofibres were more explosible than the longer fibres. The minimum ignition energies of these materials were greater than 1000 mJ and therefore the materials are not sensitive to ignition from electrostatic ignition sources.

56

Table 29: Comparison of nanopowders with micron-scale powders Nanopowder

Micron-scale powder (typical range of data)

Material

Particle size

Pmax (barg)

Equivalent KSt(bar.m/s)

Material

Particle size

Pmax (barg)

KSt (bar m/s)

Aluminium nanopowder

210 nm

12.5

449

Aluminium

Median

7-12

300-700

Median 12μm

5.2

50

Median 32μm

5.1

41

Median 160μm

0.7

2

Median 10μm

7.3

176