Printer Friendly

Case study: optimization of an industrial steam boiler system operation.


As one of the major thermal source, boilers consume a large amount of energy in industrial process and daily life. According to the report [Energy and Environmental Analysis, Inc. 2005] published in 2005, in United States, industrial boilers consumed about 37% of all energy(excluding electricity) at industrial facilities, while 28% of all energy was consumed by commercial boilers at commercial facilities. A well operated boiler system with high efficiency is essential for system performance from the energy consumption point of view.

Good news is that the development of advanced control technology makes the application of automatic control in boiler system possible, which was not even new in the 1960s for steam boilers [Hurter 1967]. Nowadays, most of the natural gas or oil fired boilers can operate with the air/fuel ratio controlled to a preset level. The control methods such as cross limiting control, oxygen ([O.sub.2]) trim control [Ronald, 1997], combined with airflow and fuel flow meters facilitate the accurate control of air/fuel ratio thus improve the efficiency and security of boiler operation.

Generally boiler efficiency is the highest at the boiler rated firing rate and decreases significantly as the boiler operated below 50% of its design load [Harrell 1997]. For industrial boilers, achieving optimal partial load operation is still a challenge while maintaining low excess air. One solution is to employ multiple boilers with well incorporated control to adapt to varied load and make the online boilers operated at higher efficiency [Harrell 1997]. Wu [2007] presented a case study of replacing the existing old boilers with new boilers combined with staging control. Meanwhile, replacing new boilers might not be cost-effective unless the boiler had reached the end use of the life. It may take 20 to 40 weeks from a drawing submittal to the final approvals regarding to the replacement, not to mention the high cost of the replacement work [Showers 2009]. Thus optimizing the existing system is more feasible and should be considered.

In practice, incomplete combustion, excessive combustion airflow, thermal loss due to high temperature flue gas, thermal loss through boiler surface, fuel type, burners, boiler load, and dirt heating surface are all the factors that influence the boiler efficiency [Kaya & Eyidogan 2010, Gupta 2011]. Once the boiler is designed with fuel type fixed, the stack loss is typically the major loss component related to boiler operation with major loss through high flue gas temperature and excess combustion air flow rate [Harrell 1997]. Therefore, optimizing the air/fuel ratio and reduce excess air, recovering the heat from flue gas are important ways to improve boiler efficiency [Kaya & Eyidogan 2010, Gupta 2011]. For boiler equipped with multiple burners, CIBO (council of industrial boiler owners) Energy Efficiency Handbook [Ronald, 1997] recommends appropriate burner use patterns with the highest efficiency should be used under partial load conditions when not all of the burners are required all the time. In fact, not all boilers with multiple burners make good use of this advantage. In this article, optimization of a boiler with two burners had been conducted with air/fuel ratio investigated in an industrial steam boiler system. The purpose of this article is to provide a successful optimization process with the improvement opportunities, which may be ignored in boiler operation.

For the study plant, as the production reduced due to the economy recession, the site boiler worked under partial loads for most of time in a year. However, the boiler has two burners which actually can be taken as two individual ones capable of operating separately. Thus, single burner mode and double burner mode were proposed and tested to extend the boiler partial load capacity, air/fuel ratio. As the operating condition changes, other methods like resetting the steam pressure lower, optimizing the feed water system were also proposed. The system performance and savings before and after the optimization was analyzed.


The boiler system of the study plant includes three 70,000lb/hr (31,751kg/hr) steam boilers with designed steam pressure of 300 psi (2,068kpa). Originally these boilers were coal fired condensing ones. Later two were retrofitted with high efficiency burners firing natural gas and oil and built with automatic control system. At the same time, the condensing system was no longer in operation due to the poor performance and high maintenance cost. Currently only the two retrofitted boilers are in operation. The produced steam is mainly used for air handing units and plant production.

Each boiler has two burners. The capacity of each burner is 400 therms/hr (11.7MW) with the low limit of 60 therms/hr (1.8MW). Each burner is configured with one constant speed 15 HP (11kW) force draft (FD) fan. The flue gas is exhausted by one fluid driven induced draft (ID) fan. The boiler system shares one feed water system including two 50 HP (37.3kW) feed water pumps driven by electrical constant speed motors.

Each boiler ran two burners at the same time to maintain the supply steam pressure set point at 110Psi (758kPa). Cross-limiting control combined with automatic O2 trim was employed for air/fuel ratio control. The boiler chamber pressure was maintained slightly negative by controlling the ID fan and ID damper. Feed water valve was modulated to maintain the boiler drum level. Feed water pumps operated continuously at constant full speed.


Theoretical air/fuel Ratio

Since natural gas is the primary fuel, the combustion analysis involved here is based on natural gas, mainly methane, which accounts for 90% more. Combustion of methane can be characterized in equation 1.

C[H.sub.4] + 2[O.sub.2] = C[O.sub.2] + 2[H.sub.2]O (1)

For complete combustion, theoretically 17.24lb (7.82kg) dry air versus 1lb (0.45kg) methane is required. For analysis convenience, this ratio is converted to 0.778kpph (kilo pounds per hour) (0.1kg/s) air versus 1kcfh (kilo cubic feet per hour) (28.3m3/hr) natural gas to match up the data provided by plant meters.

Excess Air in Practice

Practically, more air than theoretically required must be supplied for the purpose of complete combustion [Farthing 2002]. However, too much air could degrade the boiler efficiency by heating excess cold air to the high stack temperature. It could also result in the production of carbon monoxide [Wulfinghoff 1999]. If less air is supplied, carbon monoxide will accumulate, which means incomplete combustion, low combustion efficiency, and insecure operation. When the volumetric fraction of oxygen, nitrogen, and carbon monoxide were known, the excess air percentage can be found from (ASHRAE, 2001) by equation 2.

ExAir% = [100[[O.sub.2] - (CO /2)]/ 0.264[N.sub.2] - [[O.sub.2]-(CO/2)] (2)

Since nitrogen was not measured in this study, the ratio of nitrogen and oxygen volumetric fraction in ambient air yields the following equation, which was used in the calculation in this article.

ExAir% = 100[[O.sub.2]-(CO/2)]/ 20.9% - [[O.sub.2] - (CO /2)] (3)

At a certain firing rate, excess air should be optimized to a level with O2 in flue gas at the lowest possible level without CO produced or with CO level in an acceptable range.

Minimum Airflow by Code

For safe operation, NFPA 85 code requires airflow rate not to be less than the purge rate airflow, which should not be lower than 25% of full load mass airflow. For the study site boiler, at rated firing rate of 800therm/hr (23.4 MW), 4% [O.sub.2] level in flue gas was measured when complete combustion happened, considering the heating value of natural gas of 9.74 therm/kcf (36.3MJ/[m.sup.3], the purge rate mass airflow is 20kpph (2.52kg/s). So under no condition can the system air flow rate be lower than 20kpph (2.52kg/s).


Reset System Steam Supply Pressure Set-point Lower

Due to the reduced production, the boiler has been working to maintain the system steam supply set point of 110Psi (758kPa) for a long time. Since the feed water pump and the chiller driven by high-pressure steam have not been in operation for many years, the system only needs to satisfy the demand of current production and space heating. Usually steam heating system has more piping loss when operated at a higher pressure. Considering the required steam pressure for current production and heating coil are 50Psi (345kPa) and 15Psi (103kPa) respectively, the steam pressure set point was recommended to reset lower to 50Psi (345kPa) to reduce the gas consumption and system loss.

Retrofit all FD fans with VFDs

For existing control, the combustion air was supplied by constant-speed FD fans with the airflow rate modulated by the air dampers. When the standby boiler was loaded at the low limit firing rate, FD fan dampers were almost fully closed, yet 21 kpph (2.65kg/s) air leakage was measured through the dampers, which indicate a bad performance of the existing control. To achieve accurate airflow control, variable frequent drives (VFDs) were installed for each FD fan. The amount of combustion air flow was controlled by modulating the FD fan speeds while fully opening the combustion air dampers. Better control was achieved and fan power was saved.

Developing Single Burner Mode

Each boiler includes two burners capable of being operated separately. The existing system operated both burners together to maintain the boiler load. When system load decreased to a level without requiring two burners, the system was operated with one burner but the associated FD fan remaining in operation, which was not designed in the original system control. If the burner and the FD fan were both shut down, low system air flow may trip off the boiler for violating the airflow requirement by boiler code. Thus single burner mode should be tested then developed in the system control.

With new mode proposed, the original air/fuel ratio tested under double burner mode condition was no longer appropriate for use in single burner mode. Considering the control and operation upgrades made to the boiler, especially the need to build single burner mode to extend the boiler capacity to the lower load side, the air/fuel ratio under different operation conditions was re-inspected.

Inspection Air/Fuel Ratio for Single Burner Mode and Double Burner Mode

Test Procedure: Test the boiler from the lowest firing rate to the highest then back to the lowest with 10% firing rate interval. At a certain firing rate, start from the existing data in the program, use the CO level in flue gas as an indicator, if no CO was measured, reduce the air flow rate gradually till CO was measured, if CO level was too high, increase the air flow rate a little till CO level decreased. Then at the optimal [O.sub.2] level with CO at the possible lowest level, record the [O.sub.2] level, air flow rate, and firing rate.

Measurement Device: During the measurement, portable combustion analyzer from TSI with series 6213 was used to measure the flue gas. The data was compared with that measured by the site [O.sub.2] analyzer. Air flow rate and natural gas flow rate delivered to each burner were measured with Model AYR Preso Meter and EPI gas meter of the site.

Double Burner Test Result: For double burner condition, both two burners are operating at the same firing rate. After performing the test, the new [O.sub.2] trim curve was obtained which is given in Figure 1. Part of the test result is also presented in Table 1.

Data presented in Figure shows that after optimization of the air/fuel ration, the [O.sub.2] level used in the [O.sub.2] trim control was reduced, which proves higher efficiency of boiler operation at the reduced [O.sub.2] level.

Single Burner Test Result: Activate only one burner and related FD fan, keep the FD fan of the other burner off and damper closed. As required by code the lowest total airflow is 20kpph (2.57kg/s), which is the flow rate required at 140therm/hr (4.2MW) firing rate with 10% [O.sub.2] level in flue gas. When one burner fired between the low limit of 60therm/hr (1.8MW) and 140therm/hr (4.2MW), three possible ways can be used to supply the air, Case 1: supply all 20kpph air (2.57kg/s) from the operated burner; Case 2: let part of the air through the operated burner, the left through the offline burner; Case 3: make the airflow evenly supplied through two burners. The tests for all three cases were done and the results are shown in the Table 2 through table 4.

Table 2 shows that large amount of CO was produced as excessive air was supplied, which resulted in incomplete combustion and unacceptable flue gas emission level.

Comparing the data from Table 3 and Table 4, it is found that under the same load condition case 3 shows slightly higher boiler efficiency measured even the [O.sub.2] level stayed almost the same. However, 6kpph (0.78kg/s) airflow through burner 2 was too low to be accurately controlled through mechanical damper combined with slightly negative boiler pressure and the low accuracy of airflow meter at low flow conditions. Therefore, for single burner mode control, case 3 was implemented.

In single burner mode, the O2 level in flue gas was compared under two scenarios presented in Figure 2. One was to simply shut down one burner and keep the related FD fans running with the air/fuel ratio modulated based on the original double burner data, the other was the same as for the operation of the burner and FD fan but with air/fuel ratio adjusted to the optimum level. The post condition showed the excess combustion air is reduced.

Low Load Operation Limitation: It is worth noticing that actually as much as 10kpph (1.3kg/s) air flow which does nothing with combustion is supplied just to maintain the system security operation. It also indicates the low boiler efficiency under low firing rate. With regards to low load condition and minimum air flow requirement by code, one report [Shankar 1998] presented a comprehensive discussion. Though there is a demand for reducing the low load air flow for natural gas fired boilers, lack of field testing data and support from NFPA code influence each other and restrict the economical operation of gas fired boiler under low load condition. Thus, further research from both theoretical study and onsite test should be conducted.

Natural Gas Savings with Optimized Air/fuel Ratio: Based on the test result of [O.sub.2] level data, natural gas consumption under different firing rate can be estimated. It was assumed that the savings mainly came from the boiler thermal loss reduction resulted from the reduced excess air to be heated up to the stack temperature. Then the thermal loss with excess air was estimated with equation 4.

Thermalloss = [c.sub.p] x [[??].sub.ex] x {[T.sub.stack] - [T.sub.airintake]) (4)

Since the boiler system was located indoor, the intake air temperature was measured quite stable within trended two weeks, which was 72[degrees]F (22.2[degrees]C) and used in calculation. The recorded stack temperature during the test under various firing rate was used. The thermal loss difference before and after the optimization was converted to natural gas flow rate savings with its heating value. Then the natural gas savings under various firing rate (gas flow rate) was obtained and given in Figure 4 from which it was found that with optimized air/fuel ratio control, up to 7% natural gas was saved at lower load condition while savings was not significant under higher load conditions.

FD Fan Power Consumption: With the installation of VFD, the FD fan power consumption can also be estimated with combustion air required at different firing rate. The FD fan power was estimated with the curve of fan speed and air flow rate obtained during the test. The fan power was proportional to the cubic of air flow rate. The calculation sheet was given in appendix. Figure 4 presents the final saving ratio under different firing rates. It was found that with the VFD installed, more than 80% of FD fan power saving was achieved.

Feed Water System Improvement Opportunities

Feed Water System Recommendations: Feed water system works to supply water to boiler during startup, normal, and emergency operations. The system was controlled automatically to maintain the water level in the boiler drum. In the existing system, the feed water was supplied by constant speed pumps with the boiler drum level maintained by adjusting the feed water valve position. System operation data showing the feed water valve operated at lower position indicated great pressure loss through the valves. So VFDs were recommended for the two pumps with electrical motors. After reducing steam pressure from 110Psi (758kPa) to 50psi (345kPa), new control method was implemented for the feed water system. The feed water system line pressure set point was recommended to be reset from the existing fixed 110Psi (758kPa) to a lower level based on the feed water valve position. Pressure loss from the feed water valve was reduced, pump power consumption was saved.

Implementation Results: After the installation of VFD for feed water pumps and the updated system control, the following feed water system operation data are collected.


Boiler system efficiency improvement opportunities had been investigated and implemented in an industrial steam boiler system. The following energy saving opportunities were found.

1) As the production reduced, the original system operated with a high steam pressure set point, which can be reset lower according to the actual load side requirement. It reduced the system thermal loss when operated at a high system pressure.

2) At lower firing rate, the boiler was found to operate with large amount of excess air, which degraded the boiler efficiency. The boiler configured with two burners having the flexibility to adapt to varied load condition was not fully used in the existing system operation. So single burner mode was proposed. With the implementation of single burner mode combined with air/fuel ratio optimization, boiler efficiency was improved up to 7% under low load condition compared with the base case operation by simply shutting down one burner.

3) With VFDs installed on the FD fans, the amount of combustion air was more accurately controlled especially under lower firing rate, about more than 80% of the FD fan power was saved.

4) For the feed water system, with the installation of VFD and reset of feed water system pressure based on the feed water valve opening ensured the system security operation, about 78% of the feed water pump power consumption was saved. The feed water valve lifespan is extended.


Energy and Environmental Analysis, Inc. 2005. Characterization of the U.S Industrial Commercial Boiler Population. Report submitted to Oak Ridge National Laboratory.

Farthing, D. C. 2002. Combustion control strategies for single and dual element power burners. Steam Digest. Compiled for the industrial technologiesprogram. U.S. Department of Energy. Energy efficiency and renewable energy. pp29-34

Gupta, R. D., Ghai, S., Jain, A. 2011. Energy efficiency improvement strategies for industrial boilers: a case study. Journal of engineering and technology, 1(1): 52-56

Harrell, G. 2002. Steam system survey guide. Oak Ridge National Laboratory, ORNL/TM-2001/263.

Henry Manczyk, 2010. Optimal Boiler Size and Its Relationship To Seasonal Efficiency.

Hurter, A.G. 1967. Experience with automatic boiler controls. Proceedings of the South African sugar technologists' association, pp107-114

Kaya, D., Eyidogan, M. 2010. Energy Conservtion Opportunities in an Industrial Boiler System. Journal of Energy Engineering, pp18-25

Ronald, A. Z. 1997. Energy Efficiency Handbook. Council of Industrial Boiler Onwers (CIBO).

Showers, G. 2009. Industrial Boiler Replacement. Heating/Piping/Air Conditioning Engineering, January 16-21

Shankar, R. 1998. Low Load/Low Air Flow Optimum Control Applications. EPRI, Palo Alto, CA. TR-111541

Ultra-High efficiency indusrial steam generaion R&D opportunities. 2004. Results of the U.S department of energy's advanced steam generation workshop.

Wu, L., Liu, M., Wang G. 2007. CC Retrofits and optimal controls for hot water systems. Proceedings of international conference for enhanced building operations, San Francisco, California.

Wulfinghoff, D. R. Energy efficiency manual. 1999. Energy Institute Press.

Bei Zhang, PhD

Student Member ASHRAE

Yunhua Li, PhD

Student Member ASHRAE

Mingsheng Liu, PhD, PE


Bei Zhang is a project engineer in Bes-Tech Inc., Omaha, NE. Yunhua Li is a mechanical product engineer in Bes-Tech Inc.,Omaha, NE. Mingsheng Liu is president and CTO of Bes-Tech Inc., Omaha, NE.

Table 1. Double Burner Test Data

Firing Rate               Air flow (Kpph)             [O.sub.2]
(Therm/hr) (MW)           (kg/s)                      (%)

B#1          B#2          B#1           B#2

60 (1.8)     60 (1.8)     12.5 (1.6)    12.1 (1.55)   12.4
70 (2.1)     70 (2.1)     12.3 (1.58)   11.9 (1.53)   11
105 (3.45)   105 (3.45)   12 (1.54)     11.7 (1.50)   6
140 (4.2)    140 (4.2)    15 (1.9)      15 (1.93)     4.4
175 (5.25)   175 (5.25)   18.1 (2.3)    18.1 (2.33)   3.3
210 (6.3)    210 (6.3)    22 (2.83)     22 (2.83)     3.7
245 (7.35)   245 (7.35)   26.7 (3.4)    26.7 (3.4)    3.6

CO      C[O.sub.2]   NO/NOx   Tstack         Efficiency
(ppm)   (%)          (ppm)    ([degrees]F)   (%)

15      4.0          41/43    --             77.7
2       4.9          48/50    309 (154)      80.2
0       7.9          57/61    350 (177)      82.3
0       9.0          53/56    385 (196)      82.2
0       9.6          61/65    422 (217)      81.8
0       9            43/46    415 (213)      81.6
1       9.2          44/47    478 (248)      79.9

Table 2. Single Burner Test Data: Case 1

Firing Rate        Air flow (Kpph) [O.sub.2]    CO
(Therm/hr) (MW)    (kg/s)             (%)      (ppm)

B#1          B#2     B#1     B#2

105 (3.15)    0    21(2.7)    0      12.7       606
140(4.2)      0    21(2.7)    0      10.7       146

C[O.sub.2]      Tstack      Efficiency
   (%)       ([degrees]F)      (%)

   3.6         344(173)         76
   5.3         355(179)         80

Table 3. Single Burner Test Data: Case 2

Firing Rate          Air flow (Kpph)     [O.sub.2]    CO
(Therm/hr)                                  (%)      (ppm)

B#1         B#2     B#1         B#2

105(3.15)    0    15(1.93)    6(0.77)      13.6       17
140(4.2)     0    15(1.93)   6.6(0.85)     11.7        6

C[O.sub.2]      Tstack      Efficiency
   (%)       ([degrees]F)      (%)

   4.1         357(181)        77.3
   5.2         363(184)        79.4

Table 4. Single Burner Test Data: Case 3

Firing Rate           Air flow (Kpph)       [O.sub.2]    CO
(Therm/hr) (MW)                                (%)      (ppm)

B#1         B#2      B#1          B#2

60 (1.8)     0     11(1.41)    10.5(1.35)     16.9       44
70(2.1)      0    11.3(1.45)   10.2(1.31)     16.3       22
105(3.15)    0     11(1.41)    10.6(1.36)     13.5        5
140(4.2)     0    12.2(1.57)   11.3(1.45)     11.7        2

C[O.sub.2]      Tstack      Efficiency
   (%)       ([degrees]F)      (%)

   1.70        336(169)       63.50
   2.20        344(173)       67.80
   3.50        351(177)       74.40
   4.50        370(188)       77.90
COPYRIGHT 2014 American Society of Heating, Refrigerating, and Air-Conditioning Engineers, Inc. (ASHRAE)
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2014 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Zhang, Bei; Li, Yunhua; Liu, Mingsheng
Publication:ASHRAE Transactions
Article Type:Report
Date:Jul 1, 2014
Previous Article:Use calibrated whole building energy model to disaggregate retrofit savings for two medium sized office buildings.
Next Article:Annual daylight glare evaluation for typical perimeter offices: simulation models versus full-scale experiments including shading controls.

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters