Magnesium Elektron ingot production is currently based on either SF6 or SO2 as the active gas. Comparative evaluation of HFC-134a included, development of.
Magnesium Melt Protection at Magnesium Elektron Using HFC-134a 1
Paul Lyon1, Philip D. Rogers1, John F. King1, Simon P. Cashion2 and Nigel J. Ricketts3 Magnesium Elektron, Development Dept., PO Box 23, Swinton, Manchester M27 8DD England 2 Australian Magnesium Corporation, Heckerstrasse 30A 34121 Kassel, Germany 3 CAST CRC, 1 Technology Court, PO Box 883, Kenmore, QLD 4069 Australia
Abstract Several workers are actively seeking alternatives to SF6. CAST/AMC have patented the use of a HFC gas HFC-134a. This gas has a GWP 95% lower than SF6. Magnesium Elektron (MEL) and CAST have collaborated on the use of HFC-134a to achieve successful production plant trials for ingot manufacture. This paper provides details and results of those plant trials at Magnesium Elektron. Magnesium Elektron ingot production is currently based on either SF6 or SO2 as the active gas. Comparative evaluation of HFC-134a included, development of gas mixing equipment, distribution optimisation, optimisation of concentration and flow rates and assessment of breakdown products and potential implications. The results, from over 150 production-scale Mg-Al-Zn melts, demonstrated that HFC-134a could offer equal protection to SF6 or SO2. Use of HFC-134a by Magnesium Elektron for commercial production is proposed.
Introduction The use of fluorine-bearing gas atmospheres for protecting molten magnesium has been described as early as the 1930s1. However, the implementation of gas mixtures containing SF6 did not gain widespread acceptance until studies during the 1970s2. It was then realised that the use of SF6 was a major advancement in melt protection technology. Replacing salt fluxes and sulphur dioxide with SF6 resulted in improved operator health and safety and increased equipment life from reduced corrosion. Although expensive, SF6 is colourless, odourless, non-toxic, and effective at low concentrations. However, SF6 has become recognised as a very potent greenhouse gas. It has a Global Warming Potential (GWP) of 23,900 over a 100-year time horizon3 and for this reason has come under intense environmental scrutiny4,5. In response to this, the US EPA has formed a Voluntary SF6 Reduction Partnership6 and the International Magnesium Association (IMA) has assembled an SF6 Alternatives Sub-committee to actively research SF6 alternatives. The Kyoto Protocol also calls for a reduction in SF6 use. In addition to the greenhouse gas issues, sulphur hexafluoride is expensive. Several groups of researchers are actively seeking alternatives to SF6. Extensive searching of technical information by CAST researchers (Australia), yielded a number of candidate gases, subject to criteria for a suitable SF6 alternative7,8. At the completion of a series of “screening tests” involving Magnesium Elektron protection and ingot casting, one candidate gas appeared to meet all the criteria9. This gas is 1,1,1,2-tetrafluoroethane (also known as HFC-134a) and has the following characteristics:
1. 2.
It protects pure magnesium and a range of alloys It has 100-year Global Warming Potential of only 1300-1600 compared to 22,200-23900 for SF6 3 3. It has an atmospheric lifetime of 13.6 years compared to 3200 years for SF6 3 4. It has Zero Ozone Depletion Potential and is not classified as a Volatile Organocarbon (VOC)3 5. It is safe and non-toxic at room temperature 6. It is non-flammable 7. There are only minor toxic thermal decomposition products 8. It is non-corrosive at room temperature, but some corrosion can be expected at molten magnesium temperatures 9. It is approximately one half to one-third the cost of SF6 (depending upon the tonnage used and local supply conditions) 10. It is readily available worldwide from many suppliers as its main use is as a very widely used refrigerant gas. Whilst HFC-134a can be used for magnesium melt protection with a variety of diluent gases, not all diluent gases used for SF6 are suitable for use with HFC-134a in production environments. The combination of HFC-134a combined with judicious use of diluent gas is now referred to as the AMCover system of magnesium melt protection. Magnesium Elektron was assisted by CAST to develop the use of HFC-134a cover gas mixtures for production use in Magnesium Elektron’s facilities. Production scale evaluation of the new cover gas on ingot casting operations has shown that HFC-134a can be used to achieve good quality ingot, at a reduced gas cost (compared with SF6) and significant reduction in GWP. It is Magnesium Elektron’s intention to pursue this new cover gas further with the objective of introducing the gas into mainstream ingot production. This paper describes the production scale development/evaluation at Magnesium Elektron.
Production Development Magnesium Elektron currently use either SO2 or SF6 for molten metal protection. These gases are delivered in a CO2 carrier gas. The gases are used predominantly for protection during ingot casting. Protection during melting and refining is by use of proprietary fluxing techniques.
The objectives of the evaluation at Magnesium Elektron were three-fold • Determine the effectiveness of HFC-134a as a replacement for SF6 in current operations • Determine the effect of any breakdown products on personnel & equipment • Reduce cost & environmental impact These objectives were addressed initially by small-scale trials, followed by production scale evaluations carried out on over 150 melt batches. Small scale trials Melts were carried out on a 30kg scale. Variables were set within a matrix and included: • Alloys- AZ91, AM50, EZ33, Pure Magnesium • Melt temperature - 690 to 790°C • Concentration of HFC-134a (0.5 to 2 %) • Carrier gas - CO2 and dry air • Flow rates Melt protection was assessed by generating a clean melt surface & measuring time to create oxide blooms. The test was standardised for all variants. Good protection was achieved with HFC-134a in both CO2 and dry air. For a given set of melt conditions, slightly higher flow rates were necessary with dry air compared to CO2. In general it was observed that higher flow rates were needed as the melt temperature was increased. This was especially true of ZRE1 (3% Zn, 3% rare earths, 0.7% Zr, remainder Mg) where higher concentrations of HFC-134a were also required. HF measurements were made using two techniques: - Sorbent tubes of silica gel – worn by operators - Drager tubes – within 1 m of the crucible
In these open ingot-casting areas, temperatures were less than 730°C and CO2 was used as a carrier gas. In this way, any concerns regarding HF were minimised. Variables were: • Alloy type • HFC-134a concentration and flow rates Trial performance was assessed in terms of: • Internal and customer standards for oxidation and burn marks • HF and Fluoride emissions • Corrosion of steel coupons placed at locations around the casting area • Cost of gas used For operational reasons ingot production trials with HFC-134a were largely restricted to two alloys, AZ91D and AM50. This allowed continuous operation with the CO2/HFC-134a mixed gas. Working mainly with AZ91D the concentration of HFC-134a and the flow rate of the mixed gas were reduced progressively to a practical working level to cover a range of casting conditions. Ingot quality was judged to improve through out the trial period compared to standard production conditions. The surface finish was considered brighter with HFC-134a and the incidence of burn marks and oxidation was reduced. Figure 1 shows a stack of ingots produced using HFC-134a mixed with CO2. The active concentration of HFC-134a used in the cover gas mixture is now approximately half of the concentration of SF6 that used to be used at similar or lower total flows. It was observed that a thicker protective film formed using HFC134a compared to SO2 but that this was less noticeable once the concentration and flow were optimised. When flow rates, HFC134a concentration and distribution had been optimised, the ingots produced were of a quality required by exacting customer standards.
Measured HF levels outside of the crucible, in the operator environment were low & well below the UK HSE STEL* of 3ppm or TWA+ of 1.8ppm. Note that standard Magnesium Elektron fume extraction was running through out the evaluation. High levels of HF were detected inside the crucible (50ppm). Attack of steel within the crucible was significant at high temperatures (790°C) and when dry air was used as a carrier gas. Previous CAST publications have stated that HF emissions were higher using air as the diluent gas. It is possible that a relatively well-sealed lid on the crucible contributed to this by retaining breakdown product (HF) within the crucible. This would be accentuated by tests being done continuously over several days. However, these initial tests were encouraging enough to for Magnesium Elektron to pursue HFC134a gas mixtures on full-scale production trials. Production trials Evaluations were carried out in two production foundries based at Magnesium Elektron’s Manchester site. This allowed comparison in performance between SF6, SO2 and HFC-134a.
* Health & Safety Executive Short Term Exposure Limit. + Time weighted average
Figure 1 - AZ91D cast under the protection of HFC-134a Limited evaluations were also carried out on other Mg-Al, Mg-Zr and Mg-Zn alloys. Results were generally favourable. It was however noted that some improvements in gas distribution would be required to take full benefit of HFC-134a gas mixtures. HF monitoring during ingot casting was conducted using two improved (simpler) techniques:
1. 2.
Drager pocket monitor, PAC111, which gave real time analysis Silica gel impregnated filters, for subsequent analysis
Readings taken at different points around the casting area and through personal monitoring indicated that the HF levels were low. Up to 0.02 ppm HF was detected though generally levels below 0.005 ppm were obtained. These are well below the HSE TWA of 1.8 ppm. In addition to HF monitoring those operators closest to the casting area were monitored for fluoride absorption through urine tests. Selected individuals were asked to provide samples before the trials were started and then before each shift over the normal shift pattern during the evaluation period. No fluoride levels were detected beyond normal levels. Mild steel corrosion samples, placed close to and above the casting point, were used to assess the corrosive nature of the HFC-134a mixed gas and breakdown products. Through weight lost due to corrosion product it was possible to show that while there was some oxide formation this was no more severe than for the current gas systems used.
The cost figures are based upon bulk gas supply at minimum cost at the Manchester site of Magnesium Elektron. No account is taken of capital costs for mixers or licence fees. Costs will vary between applications, locations and the type of equipment used. The most important factor for cost will be how well the atmosphere above the molten magnesium is sealed from ambient air. If the above data is again used to generate information relative to SF6 it is possible to show the environmental impact of switching to HFC-134a for the type of casting operations used at Magnesium Elektron. Substituting each protective gas with its CO2 equivalent allows the GWP to be represented for the different gas mixes. See Table 2. Table 2 – Relative greenhouse gas emissions from different gas protection systems
Relative CO2 Equivalent
Gas mixing & distribution The development work carried out on HFC-134a mixtures has highlighted the need to give thorough consideration to the mixing of the gases and the distribution. Magnesium Elektron use a balanced pressure mixing system, operating at >10 bar. The low vapour pressure of HFC-134a resulted in lower mixed gas pressure and hence lower volume flow was achieved during the trials. For trial purposes, mixers were modified to accommodate the lower delivery pressure of HFC-134a for limited operations. A new mixer unit has been specified for production use, which will generate the required volume flow rates for Magnesium Elektron operations. Distribution of the protective gas will vary according to the layout of individual ingot casting machines. At Magnesium Elektron it was noted that nozzles designed to improve the flow of gas on to the ingot surface had a distinguishable effect on the effectiveness of the mixed gas. It is not clear whether this is due simply to improved control of where the protective gas is directed or whether this is associated with the lower thermal stability of HFC-134a requiring more immediate contact with the molten metal than for SF6 containing mixtures. Costs Analysis of the typical flow rates and concentrations used during the plant trials was used to provide a cost comparison with CO2/SF6 and CO2/SO2. A summary of these normalised figures is provided in Table 1. Table 1 – Relative costs for different gas mixtures CO2/SF6
CO2/SO2
CO2/HFC-134a
Flow rate
1
1.3
1
Concentration
1
0.5
0.55 to 0.78
Cost
1
0.3
0.36 to 0.43
CO2/SF6
CO2/SO2
CO2/HFC-134a
1
0.0009
0.02
It is recognised that HFC-134a is still a significant greenhouse gas. However, evaluation of different melt protection systems should be conducted with a view to total greenhouse gas emissions, not just the 100-year GWP of the active gas mixture. It should be pointed out that the amount of active fluorine-bearing gas used in magnesium melt protection is very small. The active gas is usually less than 1% of the total gas flow, which in itself is quite low. Any new technology that is capable of reducing greenhouse gas emissions by 98% should be seen as a major improvement. Particular when it is achieved with minimal capital investment and at the same time producing significant operating cost savings. The commercialisation of AMCover is continuing throughout the magnesium industry.
Use of HFC-134a gas mixtures at Magnesium Elektron Where possible, it is Magnesium Elektron’s intention to reduce SF6 useage by replacing it with HFC-134a gas mixtures. Ingot protection will be the target area for the reduction of SF6 use.
Conclusions 1.
It has been successfully shown, at Magnesium Elektron, that HFC-134a can replace SF6 and SO2 for ingot casting of a range of Mg alloys.
2.
The level of HF in the work environment during the trials was below the UK HSE TWA value of 1.8 ppm.
3.
Fluoride absorption, by operators, is not a risk under the conditions evaluated at Magnesium Elektron.
4.
The integrity of steel structures is not at risk from HFC-134a or its breakdown products under Magnesium Elektron ingot casting conditions.
5.
Greenhouse gas emissions are significantly reduced when HFC-134a is used to replace SF6.
6.
There is a significant cost benefit in replacement of SF6 with HFC-134a.
7.
Cost benefits of replacing SO2 are less attractive than for SF6, however the toxicity of SO2 makes HFC-134a a viable alternative.
References 1.
H.A. Reimers, “Method for Inhibiting the Oxidation of Readily Oxidizable metals,” US Patent, No. 1,972,317, September 4, 1934.
2.
J.W. Fruehling, “Protective Atmospheres for Molten Magnesium,” PhD thesis, University of Michigan, 1970.
3.
C. Granier and K.P. Shine, “Climate Effects of Ozone and Halocarbon Changes” Scientific Assessment of Ozone Depletion:1998, World Meteorological Organisation
4.
S.C. Erickson, J.F. King and T. Mellerud, “Conserving SF6 in Magnesium Melting Operations,” Foundry Management and Technology, Vol 126, No 6, 1998, 38-45.
5.
H. Gjestland, H. Westengen and S. Plahte, “Use of SF6 in the Magnesium Industry: An Environmental Challenge,” The 3rd International Magnesium Conference, G.W. Lorimer, Ed., The Institute of Materials, 1996, 33-41.
6.
S.C. Bartos, “EPA’s Voluntary Partnership with the Magnesium Industry for Climate Protection,” Magnesium Technology 2000, H. Kaplan, J. Hyrn and B. Clow, Eds., The Minerals, Metals and Materials Society, Warrendale, PA, USA, 2000, 83-86.
7.
S.P. Cashion, “The Use of Sulphur Hexafluoride (SF6) for Protecting Molten Magnesium,” Ph.D Thesis, University of Queensland, 1999.
8.
N.J. Ricketts and S.P. Cashion, “Hydrofluorocarbons as a Replacement for Sulphur Hexafluoride in Magnesium Processing,” Magnesium Technology 2001, J. Hyrn, Ed., The Minerals, Metals and Materials Society, Warrendale, PA, USA, 2001, 31-36.
9.
S.P. Cashion, N.J. Ricketts, M.T. Frost, and C.J. Korn, “The protection of molten magnesium and its alloys during diecasting,” Paper presented at the 8th Annual IMA Magnesium in Automotive Seminar, Aalen, Germany, 12-14 June 2000.
