Optimize Multi-Fuel Boiler Operation with Modern Control
As fuel prices rise and emissions regulations become tougher, companies must adopt new methods to remain profitable and sometimes just to survive.
Robert P. Sabin
Industrial production sites face many challenges, but for power and utility operations several challenges are especially acute, as increasing populations and per-capita energy use drive an inexorable rise in long-term energy prices. To survive, companies must adapt by addressing power and utility energy costs: they must improve efficiencies, use cheaper fuels, and eliminate waste. At the same time emissions control requirements are becoming more stringent and difficult to achieve.
This set of facts presents significant challenges for industry, yet it also presents an enormous opportunity. Consider that, according to the U.S. Energy Information Administration, industrial facilities consume as much as 50% of the world’s total energy. At the same time, according to the International Energy Agency, the industrial sector has realized only 50% of the potential system lifecycle improvements in energy. Considering that energy is typically the largest controllable cost of manufacturing that can be affected locally at a site, energy savings go directly to the bottom line. Depending on the size of a company’s operations, even a 2% improvement in power and utility efficiency can bring a $1 million benefit annually. Potentially even more significant is substituting 20% of traditional fossil fuel consumption with low-cost alternate fuel sources. This could bring an additional $2 to $3 million in annual return. Such improvement programs can be self-funding at the start, and then yield a profit via sustained benefits with additional advantage coming through improved maintenance and the avoidance of fines and shutdowns.
Penalties for CO2 production and “carbon taxes” are already in place in some parts of the world and are likely to be more widespread in the future. Figure 1 shows that, if CO2 taxes are included, shifting from petroleum or electricity to biomass can reduce fuel costs to almost zero, and going to waste gas as an energy source could actually yield an effective profit through reduced emissions penalties.
Fuel BTU variation
But optimizing a combustion process in the powerhouse to maximize efficiency and take advantage of low-cost fuels at all times is difficult, especially when attempting to operate highly reliably and responsively. Changing load in the industrial powerhouse is a challenge under the most favorable conditions, and an even larger challenge dealing with fuel variability.
Natural gas is considered a stable and consistent fuel, but even this conventional fuel can vary in BTU per volume by ±10% over time. Combustion control systems must manage even this level of variability in order to optimize energy cost results.
In an effort to restrain costs, a large percentage of industrial boilers make use of both conventional and unconventional fuels. Most often, these unconventional fuels are being produced as by-products of the main site production operations within the facility. Examples include off-gases in chemical plants and refineries, coke oven gas and blast furnace gas in steel mills, waste wood in a pulp and paper mill, or biogas at a food plant. These sources are often essentially free sources of energy for power and utility operations, but they are not typically available with consistent supply and they usually vary considerably in energy content per volume over time. The chemical composition of off-gas or waste liquids from a refinery or chemical plant can change dramatically as feedstocks and product mix change. This inconsistency is not conducive to running combustion processes in a stable manner, and therefore the use of unconventional fuel sources has traditionally been restricted. Low-cost fuels have most typically been used in a base load manner consuming a fixed volume, with traditional fossil fuels being used to achieve stability and follow load. This has often resulted in uneconomical cases where off-gas is being flared while natural gas is being used to produce steam for production needs.
The concept of fuel-to-air curves began when automated combustion control techniques were first developed using pneumatic controls. The “curve” was actually a mechanical cam in a pneumatic actuator. When a combustion process control system was commissioned, the cams in the actuators for fuel and air control were shaped such that a safe amount of air was delivered for the fuel flow over the load range. The cams in the pneumatic actuators provided a fixed air-to-fuel ratio for any particular firing rate. These implementations were simple and reliable, normally prevented an unsafe fuel rich condition, and even provided a level of excess air management with an oxygen trim function. But there was also a safety margin of additional airflow built into the implementations which resulted in combustion efficiency being compromised, and because of the magnitude of variability, oxygen trim was not fully able to optimize operations.
Pneumatic controls were eventually replaced with electronic ones, which contained circuitry that emulated the actions of the cams in pneumatic controls. These controls were somewhat more responsive and maintainable, but the control strategies that went with this technology were essentially the same as those that had been designed previously.
In the 1980s, computers entered the picture. Their developments gave the combustion process control engineer the ability to implement complex calculations and a variety of control techniques that give tighter control. But even to this day, the traditional method of implementing fuel-to-air curves for combustion control has largely survived intact.
Using this control technique that was developed 60 years ago is no longer an option. There are too many variables, and performance requirements have become much more stringent.
Too much for conventional control methods
Conventional combustion control based on air-to-fuel curves cannot handle optimization with a variety of fuel sources because it is based on fixed assumptions made at one specific point in time. As explained earlier, fuel feeds in a typical industrial site today are variable. In addition, equipment performance changes over time and ambient conditions are different every day. To avoid unsafe operating conditions, the too-common response has been simply to increase the airflow rate. These buffers are added as shown in Fig. 3 to manage variations, but they result in inefficiency and emissions. Large amounts of expensive energy are sent up the stack in the excess air while fan loads are increased and emissions equipment is overworked. But how can a combustion process with so much variability be controlled?
It can be done
In the final analysis, a boiler or heater has one control input: BTU demand. That demand may be expressed in terms of steam pressure or flow for a boiler or fluid temperature for a fired heater, but it all comes down to that one variable. It is the task of the combustion process and its control system to meet this demand optimally (which in practice translates to minimum cost and emissions) in spite of variations in fuel BTU content and rapid changes in demand. Done properly, this can turn a boiler from a cost center into a profit center, because it allows the use of fuels that would otherwise be discarded as waste or, because they are derived from renewable sources, can provide greenhouse gas offset credits.
The most advanced combustion control systems available today have the ability to infer the BTU content of fuel continuously and in real time. This allows the system to match the demand signal with a fuel firing rate signal, and calculate precisely the amount of air needed for optimal operation. In addition, these implementations allow a BTU of one fuel to be substituted easily for another, such that the use of preferred fuels can always be maximized. How do they do it?
The answer is to eliminate the use of fuel-to-air curves and move combustion control to a totally mathematical and model-based implementation. The control system should include a mathematical model of the boiler and a set of constraints using multivariable predictive control. This solution uses standard boiler instrumentation to derive a relative index of heat release in the furnace. Once this is known, specific firing rate requirements can be determined and fuel can be adjusted on a real-time basis to stabilize furnace heat release. Adjusting the incremental fuel supply along with dynamically correcting for excess air requirements results in a robust and reliable control methodology. This is much more than a refinement to existing technology; it is a quantum jump in control.
This method has been put to use in industry with good results. One example is that of a multi-fuel boiler. Prior to optimization, it suffered from large inconsistencies in steam from alternative fuel. A base load was carried on alternative fuel, while fossil fuel was used for header control. In normal operation, 60% to 70% of steam was generated with alternative fuel. But because of control limitations, the boiler was run with high excess oxygen of 8% to 10% and there were emissions issues.
Applying math- and model-based control yielded a 5% to 10% increase in steam generated on the least-cost fuel, a 1% to 2% improvement in efficiency, a significant increase in demand swing response, and improved emissions performance.
Other examples include:
- A chemical plant diverted a waste hydrogen stream to a new boiler. Optimized control maximized hydrogen use while maintaining stability. Annual natural gas use was reduced by 1 million BTU and CO2 emissions were reduced by 30%.
- A pulp and paper mill wanted to replace coal with biomass. Three boilers were retrofitted, adding a suite of advanced combustion control solutions. This resulted in a reduction in coal use of five tons-per-day and improvements in boiler turndown and powerhouse stability.
- In another situation, a food plant installed a digester to make biogas from processing waste and converted a boiler to dual-gas operation. The application of combustion optimization maximized biogas usage and reduced natural gas by 15% to 30%.
In today’s business environment, combustion processes must operate optimally at all times, despite variations in load demand, fuel BTU content, and even fuel type. By using the latest control techniques, well-operating multi-fuel boilers can often generate 90% of a plant's steam from waste and alternative fuels, operate in automatic control over 95% of the time, and maintain emissions to specified levels.
Robert P. Sabin is a consulting engineer for Emerson Process Management’s industrial energy solutions group.
For more information, visit: