In case of production lines of 250 TPD and more, heating is not needed for standard thicknesses. From the dross box until 450°C, we have to work with controlled cooling without forced convection in order to achieve the requested residual stress for a proper glass cutting and processing afterwards. The glass is able to radiate to heat exchangers under and above the glass. These heat exchangers are cooled with air in order to have large cooling range contrary to water cooled heat exchangers. The air needs to be sucked through the heat exchanger pipes, connected valves and ducts. This involves already a certain energy consumption to drive the fans.
Below 450°C, we could in principle work with only natural convection without any energy use. However, in that case the lehr should become tremendous long and so the amount of rollers and the length of the building. Therefore down to 400°C, we continue to work with radiation cooling and below 400°C, forced convection is used to shorten the lehr.
Up to last year, for a ribbon covering the full width of the heat exchangers and blowers, 10 kWh/Ton electrical energy or 25 kWh/Ton (90 MJ) primary energy was the standard for a (CNUD EFCO) lehr. With the knowledge that float glass needs about 8000 MJ/Ton, the 1% energy consumption was considered as negligible. But glass producers started to realize that with a cost of 0.1€/kWh, the energy consumption for one campaign (17 years) is of the same order as the investment cost for the lehr. Indeed for a typical 600TPD line, we get 10 x 600 x 365 x 17 x 0.1 = 3723000 €, which is of the order of the investment of a lehr. A reduction of 30 or even 50% with a small increase of the CAPEX has a nice payback and nature will be grateful.
The heat exchangers operated above 500°C are constructed from SS304 stainless steel for reasons of temperature stability and corrosion. But stainless steel is a reflector for heat radiation, which increase the throughput of air to attain a certain cooling capacity. CNUD EFCO started to coat the heat exchangers with EMISSHIELD®, a ceramic nano coating with an extreme high emissivity (absorptivity) allowing more cooling with less air.
In case one fan is used for all the control zones, we have to work with valves to regulate the cooling rate of the particular bundle. Most of the time, the valves for the edge control zones are partially closed and the center one is fully opened. The closed valve is a typical energy consumption, which can be avoided by working with separate fans for each control zone. The valve is replaced by an inverter eliminating energy consumption without any added value. CNUD EFCO calls this the multi fan solution. This can in principle be applied on radiation and forced convection zones.
For the forced convection zones, we restudied the pressure drops, caused by narrow ducts, rectangular curves, expanders …. We found that by investing in more steel and labor, a serious reduction of the pressure drop was possible in our forced convection zones (RET and F). This extra CAPEX has however a very nice payback.
For the F-zones, it is the habit to work with (rather warm) inside air above the glass. By taking air through a filter from the outside, a large gain is possible because 10°C (and more) colder air is much more efficient for the cooling. In this case, an inverter allows working with a lower energy consumption.
But also by choosing improved fans with direct drive and inverters as a standard is a good method to reduce the energy consumption when a lower load than designed is used.
All these small measures allow reducing the energy consumption with about 50% or about 2000000€ for one campaign.
More futuristic is to use the extracted heat for the production of electricity. With the proven ORC technology and liquid cooled heat exchangers, it should be possible to drive all the fans with electricity generated from the heat of the glass. But up to now, glass manufacturers are not very positive about water or thermal oil cooled heat exchangers above and under the glass. Another method should be to design Stirling motors, which are operated with the heat of the glass. Both solutions should allow working with a standard length lehr without any energy consumption. In other words, we have OPEX=0 for an acceptable increase of the CAPEX of the lehr or the produced electricity pays off the lehr.