Glass International 2011
In recent years we have witnessed increased interest in glass furnace electrical boosting systems. Is this becoming part of new business strategy in glass manufacturing?
Instead of aiming to increase furnace lifetime an alternative business case may be to produce more tons per square metre. This increased pull rate may wear out a furnace more quickly.
Our objective is therefore not to have a furnace lifetime of +12 years, but to accept a lifetime of +6 years while achieving double the throughput. In fact, this carries some further advantages such as a faster return of investment and the opportunity to innovate more often.
Such a concept faces the constraints of what the infrastructure, such as refractory material, is capable of handling when it comes to more extreme high temperatures; one of these being of course the maximum allowable crown temperature.
This is where electrical boosting comes into the equation. Of course, we can increase the amount of cullet, but for that cullet has to be available. We could also start using less energy-consuming recipes but again this will have its price.
There are also improved burner technologies to consider. In fact, electrical boosting is one of the major considerations when it comes to achieving an increase in furnace pull rate. Imagine waste heat recovery that feeds electrical furnace boosting…
There are several methods of applying electrical energy to a furnace, such as multi-tap switched or slide wire controlled transformers; both of which have their specific advantages and disadvantages. Another method makes use of silicon controlled rectifiers (SCRs) or what we call thyristors to control the power.
SCR-controlled boosting systems are considered to be the latest, solid state technology, although the principle has existed for more than 50 years (proposed by William Shockley in 1955). We have to understand that the SCR is actually just a solid state current switching device. The trick behind those SCRs is the complicated and sophisticated algorithms controlling them, which became available after the introduction of ultra-fast micro and digital signal processors.
With this technology, we are now capable of controlling SCRs so that they can adapt to all kind of situations and run these applications at the highest efficiency and best power factors possible. That is why we do not intend to fire SCRs in phase angle firing mode anymore but instead run them in burst firing mode where the specific
application allows it.
Firstly, let’s look at both phase angle and burst firing modes to understand what they mean and what advantages and disadvantages they have. Phase angle firing is the oldest method of controlling electrical power with SCRs since this firing method could be handled by analogue circuitry. It offers a very smooth control but unfortunately, it generates a lot of harmonics and normally runs at unacceptable power factors if not running above at least 70% of set-point.
An alternative method of controlling power with SCRs is known as burst firing. For this we define a duty cycle – a window of a specific amount of sine waves – in which we control single or multiple sine wave packages to control the total amount of power inside that duty cycle. The major advantage is that such a burst firing system runs at very good power factors and minimises harmonics. However, a major disadvantage with high power loads or multiple power loads is that burst firing may cause ‘flickering’.
With EPower we have a solution for the poor power factor and harmonics generation of a phase angle fired system called ‘load tap changing’. We also have a solution for the flickering disadvantage of multiple burst fired systems, which is called ‘predictive load management’.
Even a combination of both load tap changing and burst firing is possible and will normally give the best result. However, for that solution we would need to have multiple SCRs on multiple tapped transformers, thus introducing additional costs on both the power control system and on the transformers. Nevertheless, an acceptable return on investment can normally be achieved and should be always the subject of a power system’s layout consideration.
Usually, high power values are running in glass furnace boosting systems, therefore power factor, harmonics and flickering need to be considered. With simple phase angle firing, we avoid flickering issues. However, with this method it is common to run into some constraints, such as losing most of our control freedom or running the system with a bad power factor introducing harmonics. That is why Invensys-Eurotherm introduced burst firing in multiple zone boosting systems in combination with predictive load management to avoid flickering. The result of such a system is optimum power factor, minimum harmonics and, due to predictive load management – the distribution of sine wave packages over the total duty cycle – no flickering.
Recently, we installed a nine-zone furnace boosting system that runs in burst firing mode. This installation has now been running successfully for several months and demonstrating the benefits detailed above.
One of the biggest advantages of running a solid state EPower system is that it needs no maintenance and is not subject to wear. Even constantly controlling power fluctuations or active glass temperature control will not harm the system. In fact, it provides an additional controlled parameter to your glassmaking process, which will become important as soon as you consider the use of advanced process control methods.
For a better understanding we need to know that a single SCR will behave as a diode, the only difference being that it starts conducting only if the gate is triggered by a pulse. To be able to control an alternating current, two SCRs need to be in anti-parallel positions: One SCR triggering for the positive sine wave and the other triggering for the negative part of that sine wave.
To operate such a system, we need to start applying trigger signals to both SCRs using the sine wave zero crossing as a reference. For each SCR, there is a firing angle of 1800 and the longer we wait to trigger the SCR(s) the less time it will be in conduction, and consequently less power is applied to the load. Controlling the firing angles of both SCRs also controls the applied power.
There are many different considerations to be made during the design of a furnace boosting system. Initial costs, return on investment, freedom of control and mean time before failure are just a few. Using burst firing modes in combination with predictive load management can provide smooth, constant control, a good power factor and minimum harmonics at an achievable price level. Finally, it is maintenance-free.