"Changing Process Priorities when Firing Alternate Fuels"

By: Eric R. Hansen, Technical Consultant to Cadence Environmental Energy, Inc.

Abstract: A presentation on process modifications to accommodate alternate fuels and firing methods. Discussion will be made on process control strategy that makes combustion control the priority, burner modification to boost flame temperature, and the addition of energy for mixing of the combustion gasses to minimize excess air requirements. The implementation of these process modifications considers clinker quality, emissions and high alternate fuel substitution. Results from several facilities will be included. Challenges of Alternate Fuels: Many cement plants are burning alternate fuels like whole tires, tire chips, pumpable and non-pumpable hazardous waste, waste oils, consumer byproduct waste (paper diaper scraps), municipal refuse, wood byproducts and the list goes on. Each alternate fuel presents its own challenge to successfully substitute for fossil fuel without loss of quality, production or an increase in emissions. The challenges of these fuels can be generalized into these categories: Lower Flame Temperatures Difficult to Meter Variable Heating Value Difficult to Ignite Lower Flame Temperatures Many of the alternate fuels have significantly higher hydrogen content than the traditional fuel. Whether this is due to the hydrocarbon content of the fuel or the presence of free water in the fuel, hydrogen burns to water vapor and water vapor has a higher heat capacity (and volume) compared to burning of carbon to carbon dioxide. Methods must be implemented to offset the reduction of the amount of available heat at high temperature from fuels with higher hydrogen and water contents. When evaluating the substitution of fuels, one can use an Available Heating Value to compare fuels. The Available Heating Value is similar in concept to the difference between lower heating value and higher heating value but it is the heat available above 700 C. A temperature of 700 C for the evaluation since this is just above the temperature where calcination begins. The concepts of high-grade heat and low-grade heat are well discussed by Weber. Weber states, The boundary between the main and the subsidiary thermal system is assumed always to correspond to a material temperature of 550 C, since decarbonation in the rotary kiln in general already begins at this temperature. A certain amount of heat whose temperature is below the gas temperature at the commencement of decarbonation will always be left over from the calcining zone. This heat is lower-grade in the sense that, because of the low temperature, it cannot be further used for decarbonation or sintering, but only for preheating and drying. He indicates that with heat exchangers for dry raw meal, the temperature difference between the gas and raw meal can be reduced to about 150 C. See Figure 1, Temperature of the gas and material in the rotary cement kiln (diagram).

Page 1 of 12

Figure 1 Temperature of the Gas and Material in the Rotary Cement Kiln

Thus, by comparing the calculated available heat of alternate fuels and the fossil fuel, the effectiveness of the substitution and the effect on the process can be estimated.

HHV Kcal/kg Btu/lb Coal Coal Fines Petroleum Coke Liquid Hazardous Waste Derived Fuel Waste Tires Wood Hog Fuel Sawdust Municipal Waste 6920 5133 7833 6167 7889 4944 5072 3611 12456 9240 14100 11100 14200 8900 9130 6500

LHV Kcal/kg 6638 4866 7082 5396 7525 4586 4712 3174 Btu/lb 11948 8758 12747 9712 13545 8254 8481 5713

Available Heat At 700C (1292F) Kcal/kg Btu/lb 5359 3936 4889 3250 6017 3520 3693 2022 9647 7085 8801 5851 10830 6337 6647 3640

Table 1 Comparison of Available Heat at 700 C of Different Fuels

Difficult to Meter Many of the alternate fuels are difficult to meter. Whole tires, tire chips, municipal refuse and other alternate fuels are extremely difficult to meter uniformly in the time frame required for combustion. One could imagine dropping a SUV tire then one from a sub-compact car. Even with variable spacing of the tires to keep a constant tons per hour flow on a 2 or three minute average, the feed is non-uniform in the time frame of the combustion process ultimately requiring

Page 2 of 12

increased excess oxygen to maintain CO levels. Excess oxygen adversely affects the available heat from a given fuel. See Figure 2 Available Heat vs. Excess Oxygen.
Available Heat Above 700C vs Kiln Exit Oxygen

104.0% Available Energy above 1292F (700C)

4600 4400 4200 Adiabatic Flame Temperature

99.0%

4000 3800

94.0%

3600 3400 3200

89.0%

3000 2800

84.0% 0 0.5 1 1.5 2 Percent Oxygen 2.5 3 3.5 4

2600

Percent Maximum Available Energy

Flame Temperature in F

Figure 2 "Available Heat" vs. Excess Oxygen Variable Heating Value Most fuels have variable heating values, including the traditional fuel coal. Even a fuel like coal can pick up variability in heating value in the plant handling. When filling a bin with coal there is a natural segregation, the fine particles separate from the coarse particles, with the higher moisture fines feeding first and the coarse coal feeding later. Many plants can witness an obvious variability in the coal heating value even when the feed rate is uniform. Many plants experience difficulty in getting uniform feed from coal and as a result have difficulty in maintaining uniform combustion conditions. Add to this fossil fuel variability, the variability of an alternate fuel that also has non-uniform feeding characteristics, combustion control becomes quite challenging. Difficult to Ignite Many alternate fuels have different ignition characteristics than conventional fuels. Often alternate fuels have delayed ignition characteristics either due to their volatile content or particle size. Petroleum coke or tire chips may have delayed ignition and combustion rates to the point that unburned material falls onto the clinker bed and creates localized reducing conditions that effects the quality of the clinker and potentially the sulfur dioxide emissions. In actual practice, the firing of non-conventional fuels presents all these challenges simultaneously and in an additive manner. The results can be loss of production, inferior clinker quality and/or, poor kiln stability. Rather than accepting these limitations as inevitable, the challenges can be systematically addressed. The Solutions: A method of addressing the limitations of the use of alternate fuels is to compensate for the characteristics of these fuels by reducing the excess air (oxygen) to the absolute minimum the

Page 3 of 12

process requires at all times and then to reduce the minimum excess air requirement. The magnitude of the effect of excess oxygen is not often appreciated. Increasing excess oxygen dramatically reduces the amount of available heat from a fuel. increases the amount of gas volume for a given amount of available heat. Refer to figure 3, Available heat from coal vs. excess oxygen. Reducing the excess air to the absolute minimum requires characterizing the process to know what that absolute minimum is. The minimum level of excess oxygen is defined as the oxygen level required to achieve a specific emission level of CO or SO2. In some cases, the minimum excess oxygen is limited by the tendency of localized reduction of the clinker or buildups in the riser. Once the minimum oxygen level is defined, the process then needs to be aggressively controlled to this minimum. Only then, it is determined what process modifications can be made to reduce the minimum excess oxygen requirement. Step I: Establishing Excess Air Minimum The first step is to establish what the constraints are with respect to emissions. Directly affected are the SO2 and CO emissions as excess air is reduced. The SO2 emissions most often correlate with localized reducing conditions in the burning zone. One must establish what level of these emissions the process must be controlled to. This is an extremely important step. The process emissions are statistically characterized with respect to excess oxygen. Figure 3 shows a typical relationship of CO vs. Oxygen.

O2 vs CO
2000 1800 1600 1400

CO [ppm]

1200 1000 800 600 400 200 0 0.0 0.5 1.0 1.5 2.0 O2 2.5 3.0 3.5 4.0 4.5 5.0

Figure 3 CO vs. Excess Oxygen Scatter Plot, 1-minute data Figure 4 shows a relationship of SO2 vs. oxygen for a long kiln with high sulfur content in the raw materials. The SO2 emissions are a function of several factors that can vary over process type and with the characteristic of the raw materials. The sulfur to alkali balance can have a large influence in the response of SO2 to excess air. Also the amount of reduced sulfur in the raw material tends to contribute directly to SO2 emissions somewhat independent of the excess oxygen. This is particularly true for preheater kilns. In long kilns it is often observed that there is more oxygen dependence of SO2

Page 4 of 12

emissions. This is apparently due to a higher difference in the temperature of the gas vs. the raw material allowing some capture of the SO2 from the oxidation of the reduced sulfur in the raw material.

O2 vs SO2
1600 1400 1200

SO2 [ppm]

1000 800 600 400 200 0 0.0 0.5 1.0 1.5 2.0 O2 2.5 3.0 3.5 4.0 4.5 5.0

Figure 4 SO2 vs. Excess Oxygen Scatter Plot, 1-minute data Step II: Process control The oxygen needs to be controlled as close as possible to the minimum. Traditionally, oxygen control has not been very tight. Often it was determined at what minimum level that would cause operating problems and the process was controlled at some number above that. The process control priority is often the maintenance of a constant feed rate and the oxygen level was allowed to vary over some range without aggressive control. Typically the CO and SO2 measurements were not part of the control strategy. Also, NOx has become a popular process control input. Reducing excess oxygen will interfere with the value of NOx as a process parameter. This fact can be a hindrance to controlling at minimum required oxygen and hinder efforts to reduce NOx. Effective use of alternate fuels benefits from very tight combustion control. The maintenance of a constant oxygen level or a designated CO or SO2 takes precedence over the maintenance of a constant feed rate. The principal change in operating control philosophy is that if the oxygen level changes, it is probably a result of a change in the fuel. Fast action to control the fuel to keep oxygen constant will actually keep the energy (fuel) input more constant than keeping the mass flow of fuel constant. The end result the temperature remains more constant. Temperature control is achieved through feed rate control, which works well since the temperature response is slower than the oxygen response. Closed loop control of fuel by oxygen is not commonly practiced because of the serious errors that can occur due to oxygen analyzer reliability. However, todays oxygen analyzers are much more reliable and the reasonableness of the measurements can be crosschecked through CO and NOx measurements. Another step in assuring the safety of direct control of the fuel on a fast loop is to only allow the oxygen loop to control a small portion of the total coal fuel. This control approach is labeled as Fuel Trim Loop and complemented with the Low Oxygen Alarm Loop. Fuel Trim Loop A safe technique for the use of measured kiln exit oxygen in automated fuel control is the use of a fuel trim loop. In a fuel trim loop the oxygen PID is allowed to control only a portion of the fuel, say up to 1 TPH of a 10 TPH firing rate. This control loop trims 0-1.0 TPH off the fuel setpoint

Page 5 of 12

entered by the operator into the master controller in order to control the kiln exit oxygen at the setpoint. For example, if the operator has a 10.0 TPH setpoint for total coal and a setpoint of 1.2% kiln exit oxygen, this control loop will vary the coal rate between 9.0 and 10.0 TPH to keep the oxygen at the setpoint. If the setpoint were not achieved at either 10.0 TPH or 9.0 TPH, the operator would have to manually change the master controller so that the actual fuel rate is floating in the 1 TPH range of the fuel trim controller. The fuel trim control strategy is used for safety reasons. It limits the control range of the oxygen measurement that at times may be unreliable. It also never puts on more fuel than the operator sets on the master controller. This prevents excessive fuel when the oxygen may be high due to other reasons. The use of the oxygen trim loop is particularly effective when variable fuels are used. It allows fast tracking to keep fuel input more constant when the mid-kiln fuel stream is variable in weight or composition. In decision table control systems, like Linkman, in which the decision table is too slow, the analog loop is much faster in tracking the mid-kiln fuel rate. The decision table control system should make decisions on the excess O2 level and only adjust the master fuel rate when the trim loop is either at maximum or minimum. Direct control of the fuel rate by a decision tree is too slow. A decision can also be made on the excess oxygen level based on CO. Optimum excess air for maximum thermal efficiency is when CO just starts being formed. Such a control strategy will always seek a little CO. This will also result in a dramatic drop in NO, particularly with mid-kiln firing. CO has also been used directly instead of oxygen in the fuel trim loop. After some experience with the kiln exit oxygen controlling the fuel trim loop, the use of CO as the input signal (inverse action to O2 of course) may be an improvement. For a precalciner, where the calciner fuel is controlled from the calciner temperature, the oxygen control loop trims feed from the precalciner. When the oxygen goes below set point, feed is trimmed from the precalciner. The temperature increases and as a result the temperature control loop fuel is reduced, increasing oxygen. Again, the range of the trim is limited, perhaps 3 to 5% of the maximum feed, to limit a potential upset to the process in the case of inaccurate oxygen readings. Low Oxygen Alarm Loop, or Fast Acting Loop A very useful loop, particularly during initiating secondary firing and when the solid fuel has variable combustion characteristics, is the low oxygen alarm loop. This loop will cause skipping the introduction of solid fuel (or a step reduction in the primary fuel rate) when the kiln exit oxygen is below a predetermined level. The control strategy uses a value below the oxygen controller setpoint as an off-line limit. A second valve below the oxygen setpoint is used as an on-line limit. For example, with an oxygen setpoint of 1.5%, an off-line value of 0.8% and an on-line limit of 1.0% are entered. With an oxygen setpoint of 1.5% on the fuel trim controller, the following would occur. When the kiln oxygen reaches 0.8%, the solid fuel would stop being introduced (or a step reduction of 0.25 tph of fuel). As a result of the fuel rate reduction, the kiln exit oxygen should start increasing. At 1.0% oxygen, the mid-kiln fuel addition restarts (or the step change is added back). Simultaneously, the fuel trim loop should have been ramping back the primary fuel setpoint so that tendency to reach 0.8% oxygen should be reduced in the next cycle. This control strategy is very powerful in preventing CO spikes particularly for fuels like whole tires and bales. These spikes usually occur when a fuel module is introduced and there is not adequate oxygen. If there is not enough oxygen to burn the fuel, its not introduced, preventing the spike. This control strategy gives the analog controller or operator time to respond to the need of reducing the fuel rate setpoint.

Page 6 of 12

These two control loops work together. The oxygen PID loop works with a modest speed and the Low Oxygen Alarm Loop acts fast when appropriate. The incentive to minimize oxygen is high. It has been measured on a precalciner kiln that a 1.0% reduction in average oxygen will allow a 5% increase in production at the same exit gas flow rate.

Figure 5 Kiln Production vs. Excess Oxygen The objective of the oxygen control strategies is to control the excess oxygen as close to the minimum as possible. The reduction in excess air will offset potentially higher combustion gas volumes from alternate fuels with higher hydrogen and water content for the same amount of available heat. Step III: Process Modification After the process emissions have been characterized and process controls have been defined to make combustion control the priority strategy, then the actual process must be examined to determine if there are opportunities to reduce the minimum excess air requirements. The most common areas of opportunity are in addressing Fuel Ignition & Sulfur Cycle Fuel Flow Uniformity Stratification Fuel Ignition & Sulfur Cycle A limitation to the reduction of excess air is the avoidance of localized reducing in the kiln, which affects clinker quality. This can be identified by the presence of yellow centered clinker and often-increased SO2 emissions. This can be addressed through better positioning of the burner, the use of a compact, nondivergent flame and improving the ignition of the flame. Avoidance of locally reducing conditions requires that flame impingement on the clinker bed be avoided. Directing the flame on the kiln centerline parallel with the kiln axis will keep the flame away from the clinker bed and the walls. Compared to a straight blowpipe burner, a pipe equipped with a bluff body produces a more compact flame and tends to give more stable ignition. The more compact flame is easier to position in the kiln. Another factor that has a significant benefit is the stabilization of ignition. The longer the pre-ignition plume is on the flame, the more secondary air enters the flame body before ignition occurs. As a result, the mass of the flame (fuel and combustion air) increases faster than the coal particles can burn. As a result, there is a cooling of the flame and the highest heat flux from the flame is

Page 7 of 12

quite uphill of the burner. This requires that the sintering zone be held quite high in the kiln. The addition of a stabilizing torch (or other means of stabilizing ignition like a bluff body) will bring the maximum heat flux closer to the burner tip. Since more coal is being burned before the secondary air increases the flame mass, the flame is hotter closer to the burner tip. This allows the burning zone to move from uphill of the flame to under the flame. In this position of the kiln oxygen available to the clinker bed is higher. The observation is that this profile reduces the tendency for localized reduction and allows the operation of the kiln at lower oxygen levels without adverse effect on clinker quality, SO2 emissions and build-up. The stabilized ignition achieves this through quicker burnout of the fuel or the presence of increased oxygen under the flame. Figure 6, Bluff body burner for stabilized ignition.

Figure 6 Bluff Body Burner The attention to the burner can offset other adverse effects of alternate fuels. Some fuels are difficult to comminute. When introduced into the front of a kiln the fuel can fall into the clinker bed causing localized reduction. Attempts can be made to try to inject them further up the kiln. An alternative is to increase the heat flux from the flame and move the burning downhill into a more oxidizing region. When secondary firing is used like tires on the feed shelf, flame cooling due to high excess air can be offset by focusing on improving the ignition to get more combustion done before the secondary air can mix into the flame. On a practical sense, for kilns using secondary firing that increases the excess air in the main flame, indirect fired kilns achieve higher substitution rates due to the ability to maintain a higher flame temperature. Conversion of a direct fired kiln to semi-indirect firing will allow higher secondary firing rates since flame temperature throughout the entire flame length can be kept higher. It is not the added heat recovery from the cooler for indirect firing that has the biggest influence on the flame characteristic; it is that the flame is kept closer to its stoichiometric ratio throughout its length increasing its temperature. With 40% primary air, the flame is quite cool when the coal particles are 10% combusted. A flame with 12% primary air is a lot hotter at 10% completion of combustion. Thus, semidirect firing can significantly increase the heat flux of the flame in the first few meters off the burner tip. The key to achieving high fuel

Page 8 of 12

substitution rates is paying attention to the burner to maintain the flame temperature as high as possible. Getting fast and stable ignition is a key step. Fuel Flow Uniformity Improving the fuel flow uniformity can reduce the excess air requirement. A component of this is achieving as uniform fuel flow as on the time frames that are important to combustion. The gas residence time in a long kiln is about 15 seconds. The effective time with respect to combustion is only a fraction of this. Fuel flow uniformity is measured on the time basis of fractions of a second. Variability causes the average oxygen level for a given CO or SO2 to increase. Fuel flow uniformity of both the alternate fuel and primary fuel must be considered. Often improvements in primary fuel handling system can increase the ability to substitute alternate fuels. Because of the attention being drawn to the need to improve fuel flow uniformity, new technologies are being provided to the industry. See figure 7, Example of fuel metering technology to improve dosing. For a given investment, improvements in the primary fuel system should not be overlooked as a means to improve overall fuel flow uniformity. Addressing items like covered coal storage to minimize coal-handling difficulties may actually result in increased secondary fuel use. (Lower moisture also helps the flame temperature.)

Figure 7 Example of Fuel Metering Technology to Improve Dosing

Stratification Addressing stratification in the process is a means of reducing the excess air requirement. Figure 8 shows the effect on excess oxygen requirement as the result of a process modification. The process modification was the introducing energy into the middle of a kiln to reduce stratification and increased secondary fuel firing. Figure 9 shows the corresponding effect on the SO2. As a result of minimizing stratification the excess air could be reduced by a full percentage point with no increase in CO or SO2 emissions. This excess air reduction allowed the secondary fuel substitution rate to be increased up to 35% while maintaining satisfactory flame temperature. The reduction in excess air requirement allows a reduction in the ID fan flow and the potential benefits of reduced dust loss and efficiency improvement.

Page 9 of 12

O2 vs CO
2000 1800 1600 1400 1200 1000 800 600 400 200 0 0.0 0.5 1.0 1.5 2.0 O2 2.5 3.0 3.5 4.0 4.5 5.0

CO [ppm]

baseline

With High Pressure Air Injection

Figure 8 Shift in oxygen requirement for CO control with the injection of high-pressure air to eliminate stratification.

O2 vs SO2
1600 1400 1200

SO2 [ppm]

1000 800 600 400 200 0 0.0 0.5 1.0 1.5 2.0 O2 2.5 3.0 3.5 4.0 4.5 5.0

Baseline

With High Presssure Air Injection

Figure 9 Shift in oxygen requirement for SO2 control with the injection of high-pressure air to eliminate stratification. NOTE: Referenced baseline for figures 8 and 9 was with 10% mid-kiln fuel substitution; after modification data is with 20% mid-kiln fuel substitution (two tires per revolution).

Page 10 of 12

The use of energy injection also has the potential to change the CO to oxygen relationship in precalciner kilns. Modern Computational Fluid Dynamic (CFD) modeling has the ability to predict the opportunity to reduce the excess oxygen required for a given CO in a precalciner kiln. Likewise, the injection of energy to cross sectional mix the exit gasses from a precalciner kiln enables the use of feed shelf firing of whole tires. Without the mixing, the firing of tires caused unacceptable buildup in the riser. Now with a cross sectional mixed kiln exit gas entering the riser, buildup is not a problem enabling the firing of whole tires.

Figure 10 PreCalciner Kiln Modified with High Pressure Air Injection to Reduce Stratification Conclusion: The use of alternate fuels provides the incentive to reevaluate process priorities. Combustion control being prioritized over constant feed can compensate for the inherent characteristics of non-conventional fuels. The key item one has control over is the excess oxygen. The process needs to be characterized to determine the minimum excess oxygen required to obtain the emission and quality requirements and the process should to be controlled as close as possible to this minimum. Then the process needs to be evaluated and modified if necessary to reduce the minimum oxygen requirement. Attention must be placed on uniformity of fuel flow, including the primary fuel. Attention to ignition of the fuels and flame shape can shift the emission relationships and oxygen requirement. Further, the relationship of SO2 and CO emissions to oxygen can be shifted to lower excess oxygen levels by addressing gas stratification in the process.

Page 11 of 12

References: Weber, Dr. Paul, Heat Transfer in Rotary Kiln with due Regard to Cyclic Processes and Phase Formation, Bauverlag GmbH Wiesbaden, 1963 Hansen, E.H., The Use of CO and Other Trace Gasses for Process Control, IEEE Transaction on Industry Application, Vol. IA-22, NO. 2, March/April, 1986 Hansen, E.H., Staged Combustion for NOx Reduction Using High Pressure Air Injection, 44 Annual IEEE Cement Industry Technical Conference, Jacksonville, FL, May 2002
th

Figures: 1. 2. 3. 4. 5. 6. 7. 8. Temperature of the Gas and Material in the Rotary Cement Kiln Available Heat vs. Excess Oxygen CO vs. Excess Oxygen Scatter Plot, 1-minute data SO2 vs. Excess Oxygen Scatter Plot, 1-minute data Kiln Production vs. Excess Oxygen Bluff Body Burner Example of Fuel Metering Technology to Improve Dosing Shift in oxygen requirement for CO control with the injection of high-pressure air to eliminate stratification 9. Shift in oxygen requirement for SO2 control with the injection of high pressure air to eliminate stratification 10. PreCalciner kiln modified with high pressure air injection to reduce stratification

Table: 1. Comparison of Available Heat at 700 C of Different Fuels

Page 12 of 12