Glass melting using concentrated solar heat

• Main Author: Ahmad, Syed Qasid Safeer
• Other Authors: Rothman, Rachael ; Hand, Russell J.
• Published: University of Sheffield 2017
• Subjects: 621.47
• Online Access: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.707121

Glass production is an energy intensive process, primarily requiring high temperature (≈1500oC) heat, usually provided by combustion of fossil fuels. Concerns about the future depletion of fossil fuels and the emissions due to their combustion have driven a shift towards the use of alternative energy sources. Concentrated solar radiation is the only form of renewable energy which could directly provide the high temperature process heat required for glass production, without the otherwise necessary intermediate conversion to electricity and thereby avoiding the associated efficiency and capital costs. The technology for concentrating solar radiation using fields of heliostat mirrors to collectively generate a high intensity beam at a central focal point, is already well developed. However, the ability to use of such a solar beam to effect a viable glass making process involves significant challenges associated with the fact that glass furnaces typically require continuous and consistent process heat smoothly distributed, over large areas while concentrated solar radiation is only intermittently available and provides heat over relatively small areas with steep intensity gradients. In this project, experiments were conducted to incrementally develop and investigate the feasibility of an efficient and scalable process for manufacturing useful glasses with the primary process heat demand provided by a realistic concentrated solar beam. A High Flux Solar Simulator (HFSS) consisting of an array of xenon arc lamps, each coupled with an ellipsoidal reflector, was used to generate an artificial, controllable ‘solar’ beam. First, the HFSS beam was used to directly irradiate pure silica and soda-lime-silica glass forming batches contained inside insulated, refractory crucibles and directly irradiated by a realistic solar beam (intensity < 2000 kW/m2). All silicate batches generated complete melt pools with full conversion from crystalline raw materials, to x-ray amorphous glasses but only soda-lime-silica batches would generate melts with sufficient fluidity to homogenise and remove entrained gases (fining) and thereby produce transparent glasses. Next, a series of experiments were conducted to investigate factors affecting the continuity of the process. Glass forming batch, directly irradiated by a vertical HFSS beam, was contained inside a modified apparatus which enabled the melt produced to sustain a continuous flow . A specially made batch feeder enabled additional raw materials to be intermittently fed into the melting zone while the beam was still on. It was demonstrated that a semi-continuous glass melting process could be realised using primarily concentrated solar heat. However, secondary heating was necessary to sustain the flow beyond the focal spot of the beam while avoiding fracture of the crucible due to thermal gradients induced by the beam. Also, the efficiency of this process was very poor (3-4%) and the throughput was very low. This was mainly due small size of the melt limited by the area of focal spot (~6 cm dia) of the beam and the radiative heat losses from the exposed surface of the glass forming melt. Finally, a scaled-up solar-heated glass furnace was designed, built and tested, to address the remaining issues of efficiency, throughput, secondary heating demands, intermittent solar radiation availability and glass quality. The furnace was essentially an insulated box, containing the glass forming melt with an aperture in the roof for the vertical HFSS beam and raw material inlet. Integrated electrical resistance heating elements provided the secondary heating required to sustain continuity of the process by minimising thermal gradients during periods were solar radiation was unavailable. The HFSS beam converged at the inlet aperture and then diverged inside the insulated cavity to irradiate the surface of the glass forming melt. This provided a much larger surface area for glass melting relative to the focal spot of the beam. This resulted in both greatly reducing re-radiation heat losses and increasing the productivity compared to the previous experiments in which the beam was directly focussed at the glass melting surface. Also, a specially made flow control system was developed, enabling the glass melt to accumulate in the crucible until fully melted and fined before and then extracted to demonstrated production of pressed glassware. With 5.26 kW of radiation from the HFSS beam entering the beam inlet aperture of this solar-heated glass furnace, 300 g of soda-lime-silica glass forming batch was periodically fed, requiring 16 minutes between consecutive feeding cycles required to fully melt the batch and recover the glass melt temperature to 1460oC, which corresponded to a thermal efficiency of 16 %. It was shown that complete melting and fining of melt pools with surface areas an order of magnitude larger than the focal point of a solar beam could be sustained without any secondary heating. Also, beam power was switched off intermittently during melting and overnight which simulated realistic operation as per natural solar radiation availability and during these periods, the secondary electrical heating automatically provided sufficient power to avoid thermal shock. Analysis of the performance of the scaled-up solar-heated glass furnace suggested that it could not directly compete with conventional large scale (~300 tonnes/day), continuous glass tank furnaces. However, in markets where conventional glass manufacturing is infeasible due to insufficient local demand, a solar-heated glass furnace appears more commercially attractive to meet smaller local demands. For example, it is estimated, that with realistically achievable improvements expected on commercial scale up (30% glass melting efficiency), a 4 tonnes/day solar-heated glass furnace, requiring a total initial capital investment of $13M and a land area of 2 hectares, would have a payback period of 6 years for oil at 50-60 $/barrel.