- THE MAGAZINE
The part of the DCS approach to process control that made these systems so essential is what is now referred to as the human machine interface (HMI), which is the series of display screens the operator uses to control the process. These screens depict the process in a way that enhances the operator’s ability to comprehend what is going on. By far, the most important part of specifying an HMI is the ability to store historical data—essentially tape recording what the process has been doing and playing it back in a form that shows exactly how the process performs over a period of time. To fully grasp what any given process is doing, it is necessary to first be able to gather this information accurately, and then be able to view this data covering a day, a week, a month, or perhaps even a year at a time.
These abilities are an essential part of understanding—and controlling—a system as incredibly interactive, complex and delicate as a ceramic tunnel kiln. The benefits of accurate instrumentation and a good HMI with historical data archiving/display abilities include significant energy savings and product quality/consistency improvements. The payback for a properly installed and configured controls integration program can occur in a matter of months. Even with such a rapid payback, it is what the system does for the kilns’ long-term operation that makes it such a good investment.
Many companies have begun investing in computer controls—but fancy computers with high-priced software aren’t worth much if they are not used properly. The purpose of the computer isn’t to control the process, but to develop an understanding of how the process functions. Accurate measurement tools and an HMI are needed to truly maximize the firing process. The measurement tools collect the data, and the HMI communicates with the controls (stand-alone controllers, PLCs, motor drives, sensors, etc.) and provides the essential information that makes it possible to develop synergistic operation of all of these components.
Controlling the VariablesA typical tunnel kiln generally has at least several dozen individual controllers—some to control temperature, some to control pressure, and others to control flow. But it’s important to remember that these temperatures, pressures and flows are all connected to each other by virtue of the fact that all of the piping is connected at the kiln. The slightest change in one of these variables (such as the flow of air or natural gas) creates disruptions in all of the other parts of the system.
Paying close attention to the flows, pressures and temperatures is part of operating any furnace. In many cases, kiln operators try to carefully control the temperatures and pay little attention to the flows and pressures. However, the key to achieving efficiencies in furnace management is establishing a smooth and synergistic interaction among all of the controlled parameters. An HMI can help kiln operators establish this interaction by tracking and providing valuable information about each of the variables. But before they can truly benefit from the advantages of an HMI, kiln operators must first learn how to interpret that data and use it to their advantage.
TemperatureAt first glance, temperature control appears to be one of the more important issues in the typical ceramic process. In fact, many companies regularly struggle to keep the temperature indicated on their controllers within ±2ºF. But while temperature control is important, it is not nearly as critical as many other variables for several reasons. While thermocouples are tremendously reliable instruments, they really aren’t very accurate. For example, in a kiln operating at 2000ºF, 2ºF reflects a mere .05% of this span. When this number is placed alongside the numbers that reflect the load that is being heated (keep in mind that this is the mass of the kiln refractories, as well as the product) it quickly begins to look unattainable. For instance, the speed of the load moving through the kiln is rarely controlled to .05%, and this sort of accuracy is not typically expected from the typical kiln propeller system. Additionally, thermocouples are usually hiding in a protective tube—at least several inches of hot gases, a layer of ceramic tube and generally a small volume of somewhat cooler air inside the protection tube insulate them from the load. And what about the various cold junctions that end up in a typical system? Have these been accounted for in the quest to maintain the control zone temperatures to within .05%?
Junction rot is another issue to consider. Many thermocouples degrade over time as a result of exposure to high temperatures and oxygen. Some are placed in cooler environments and are exposed to condensing vapors that are often corrosive. Are these checked on a regular basis? Are the controllers calibrated to the actual thermocouple that they are wired to? If problems occurred with the thermocouples themselves, would they be spotted? Let’s say that as a result of junction rot, the thermocouple is slowly producing less voltage at the same temperature. If the signal is actually .1% low, it is impossible to expect the load to be controlled within .05%. In fact, unless all of the thermocouples in a given kiln have been replaced within the last few weeks, this level of accuracy is out of the question.
In a tunnel kiln, the load experiences heat transfer in two distinctive modes—convective and radiant—at varying levels throughout the length of the kiln. Additionally, the distances from the thermocouple to the load vary considerably as the cars pass by. As a result, any changes in the flow of hot gases in convective mode, or the slightest irregularity in the surfaces that radiates to the thermocouple in radiant mode, will show up as undulations in the indicated temperature. Space in the load between cars can also be inaccurately quantified by the controls. The occasional thermocouple car passed through the kiln will provide an idea of what the load actually experiences as it goes through the kiln, as well as some information that can be used to establish the temperature setpoints—but it offers nothing to quantify these undulations in temperature. Unless this information is processed and archived in a way that permits direct comparison with the kiln controls, such as with good HMI software, it will not be very useful to the kiln operator. Additionally, the same issues that affect the control thermocouples—cold junctions, and the vagaries of thermocouples in general—apply to these reference thermocouples as well.
PressurePressure, on the other hand, is an extremely important parameter in controlling any combustion process. Drafting of the products of combustion (POCs) is probably one of the least understood parts of kiln control, judging from the equipment that is typically seen in this extremely delicate operation. Particularly important is the ability to sense the draft suction accurately enough, and with enough resolution, to move the POCs through the system at a precise rate that perfectly matches the rate at which they are put in. This has to be done reliably on a continuous basis without constant intervention by the operators and technical staff.
While many kiln operators think that controlling the pressure to within a few tenths of an inch water column is pretty accurate, it can—and should—be much more tightly controlled. Even a few hundredths of an inch is pretty coarse for this application, particularly in tunnel kilns. Accuracy to a few thousandths of an inch starts to make a difference in the stability of the system, and the technology exists to successfully control pressure to within a few ten thousands of an inch using some of today’s more advanced sensors. The benefits in energy savings and increased product consistency are often significant.
At Kentwood Brick, for instance, digital communication cards were installed in each existing Honeywell DR4500 circular chart recorder/controller as part of the kiln updates. These cards continuously communicate with the computer, which gathers and sorts data that can be viewed through the HMI at any given time. The HMI provides “real time” numbers on what each controller is doing, while the chart depicts what it has done over, say, the last 12 hours. In “real time,” the actual kiln exhaust pressure is well within .001 in. water column (w.c.). The pressure is routinely returned to within a few ten-thousandths of an inch w.c. each time the entrance door is opened to insert a new car.
This information is essential to properly controlling any combustion process, and it becomes even more critical when dealing with the kiln exhaust system. Before Kentwood Brick updated its kiln, operators regularly made pressure setpoint changes of 0.2 in. w.c. As a result, the kiln was in a constant state of upset—essentially never able to achieve equilibrium in the critical flow of exhaust gases and cooling air. By more carefully controlling the kiln exhaust pressures, the company has been able to drastically improve its ability not only to control these flows, but also the ability of the entire system to maintain the setpoint temperatures. As the overall stability of the system improved, the energy input required to maintain equilibrium dropped dramatically.
Controlling pressures in any continuous process is a matter of controlling flow. To establish the pressure within a relatively loose box (if not open, as is the case in many kilns) the kiln operator must establish a relationship between the flow of fluids forced in and those being removed. More going in and less going out will create a positive pressure, while less going in and more going out will create a negative pressure. Establishing the right relationship takes a great deal of time and experimentation. If accurate measurement instrumentation is being used, a properly configured HMI historical data display can help immensely in determining what works and what doesn’t, and also makes it possible to maintain the right levels.
FlowCertain situations that are currently controlled primarily with pressure really should be controlled based on flow. Cooling systems, for example, typically force ambient air of unknown density and moisture content into the load, and these factors seriously affect the ability of any heat transfer media to transfer heat. And what about those pressure control loops whose operating characteristics change from night to day? Is some complex algorithm used to quantify the differences between night air to day air? Is this based on light level or the time of day? And how is Daylight Savings Time factored in?
Once again, sensors become exceedingly important. They must have fairly good accuracy and resolution, be rugged and reliable, and, in most applications, must impart a negligible pressure drop, or DP. It also helps if they are not particularly difficult to install and can be easily removed for inspection, cleaning and repair/replacement.
Many improvements have recently been made in the equipment available for measuring flow. Orifice plates or instruments that use the principles of differential thermometry to determine mass flow (velocity) times the duct area generally work the best in ceramic manufacturing.
In addition to being accurately measured, the flow must also be accurately controlled. For this operation, there is no substitute for applying a good variable frequency drive (VFD) to the fans to move the fluids. It is simply not possible to get the required resolution from a damper and control motor or from manual control.
Mass FlowControlling mass flow is also extremely important for properly controlling many combustion processes. Mass flow entails sensing flow in a manner that removes the very significant influence that temperature has on the density of the various fluids that are now very precisely moving through the system. For example, in a combustion system that recovers waste heat through the use of preheated combustion air, very accurately quantifying the mass flow allows the fuel/air ratios to be controlled so precisely that significant improvements in efficiency are generated. In a typical high-input system, such as a steel furnace requiring 300-500 million BTUs/hr, energy savings as high as 35% are routinely generated. When installed correctly, the payback on this type of mass flow system is amazingly fast—and the benefits continue for the lifetime of the system.
At Lukens Steel in Conshohocken, Pa., for instance, a new slab heating furnace was installed. As is fairly typical, the combustion system used combustion air that was preheated by the exhaust gases. The furnace was not achieving anything close to the designed heating rates or specific fuel consumption. Although all of the elements of mass flow control were in place (in this case orifice plates and thermocouples in the combustion air supply system), they were not being used correctly. Flow control was extremely erratic, and control of the fuel-to-air ratio, and therefore the furnace atmosphere, suffered as a result.
Since many control zones were being supplied by a single huge blower, the application of a VFD was impractical. Instead, careful linearization of the control valves and actuators was needed. Once these areas were addressed, the fuel-to-air ratio problems disappeared, fuel consumption dropped to 65% of where it was previously, and product throughput more than tripled. The same concepts easily carry over to the operation of any continuous or periodic kiln.
Mass Flow and PhysicsWhile all of the aforementioned principles can be applied to all of the flows in the kiln system, they become extraordinarily important when controlling the cooling section of a tunnel kiln. This is because when heating or cooling any load, we must deal with some pretty hard and fast rules of physics and heat transfer. It takes a very mathematically predictable mass of cold air at a very predictable DT to bring a fixed load mass down in temperature at a given rate. Likewise, the mass and DT of hot gases required to bring a load up in temperature at a given rate can also be accurately predicted. Pressure has absolutely nothing to do with it—unless the fact that the cooling air at a given pressure has a certain density is considered. Yet many companies continue to try to control flow-related parameters in the cooling section of the kiln based on pressure, primarily because pressure sensors cost much less than accurate, reliable mass flow sensors. In many cases, however, 10 to 15% energy savings can be achieved just by replacing the pressure sensor with a mass flow sensor—paying for the more expensive instrument in a few days.
At Kentwood Brick, for example, the supply of cooling air was controlled conventionally by reference to pressure. Even though the controller was able to maintain the pressure on setpoint with no problem, atmospheric changes (temperature and humidity) that are part of operating a furnace in an open building appeared to create wide swings in the operation from night to day. Within hours of implementing a mass flow control strategy for the cooling system, fuel consumption dropped by approximately 15% with no change in kiln speed, and the variances in operation are a very small fraction of what they once where (although not entirely eliminated).
Provided that the cooling system ductwork is laid out in a manner that permits the required upstream and downstream lengths of undisturbed flow, installing a thermal dispersion type mass flow sensor is remarkably simple. A hole is drilled, a sealing gland is mounted, and then a probe inserted. This simple tool will enable the kiln operator to learn a great deal about the process. Often, companies will learn that their blowers are oversized and that their control loops are much more stable than they were before. In many cases, control loops that were being operated manually can now be placed in automatic control. The neutral zone will now be solidly located in the kiln, and it can be moved very precisely to where it is needed. The fuel flows might also drop to levels that were never seen before—and when this happens, the payback time can be precisely calculated.
The Tools of the TradeComputers integrated with control systems have become a way of life in industry. But installing a computerized system is only the first step—a truly optimized firing process also requires accurate measurement tools and a reliable HMI with data tracking capabilities, as well as a kiln operator who understands how to read and use the information provided by the HMI to its maximum advantage.
High-tech pressure sensors, accurate mass flow meters and HMI systems can be expensive—but the cost of not controlling the firing process is often far greater than the cost of installing the right equipment. Additionally, these systems can have lasting benefits on a company’s energy efficiency and product quality, often paying for themselves within just a few years and saving hundreds of thousands of dollars over the long term.
Understanding the information provided by these tools is also key. By installing the right measurement tools and learning to read and understand the information provided by the HMI, Kentwood Brick went from making pressure setpoint changes of 0.2 in. water column to being squeamish about making changes of .005 in. water column. It went from having no idea if its pressure sensors were accurately zeroed (or were accurate at all, for that matter) to checking its pressure sensor daily and complaining that sometimes the sensor is off by .002 in. water column. It went from worrying about temperature control to within ±2ºF to realizing that a few degrees’ swing at 2200ºF is likely to be caused by the load passing under the thermocouple. It went from trying different setpoints in the morning and evening to reduce the effect of environmental changes, to knowing that these changes in conditions create large changes in corrected flow—and compensating for these changes through the application of appropriate sensing technology made the entire kiln more stable. It went from knowing only what the controls were doing at a given time to knowing exactly what the controls have been doing for the last year. And it went from not knowing how the various controls affected each other to being able to click on a button and see exactly what effect certain variables have on other variables. As a result of all of these changes, the company significantly reduced the amount of energy required to produce its product and also increased its product quality.
The technology exists today to make this happen for any brick or ceramic manufacturer—but you have to be willing to invest in it and learn exactly how to apply it before you can expect any real benefit to emerge.