Press Room

Download the PDF version.

K-Fuel Combustion Demonstration At Black Hills Power, Neil Simpson Unit I

Prepared for:
KFx Inc.
55 Madison Street
Denver, CO 80206

Prepared by:
Quinapoxet Solutions
Stephen A Johnson
30 Hickory Lane
Windham, NH 03087

Released: May 26, 2006
Reference Report: 5019-ENG-19

EXECUTIVE SUMMARY
The K-Fuel™ (K-Fuel) production plant located outside of Gillette, Wyoming started production in late 2005 at which time KFx began seeking utility and industrial boiler sites where they could demonstrate the potential advantages of K-Fuel as compared to the coals normally burned at each site. It was logical then that the first utility test site would be Black Hills Power, Neil Simpson I, because this boiler is within trucking distance of the production plant and small enough to make this test economical.

Soon after KFx reached a tentative agreement with Black Hills Power for this test, Quinapoxet Solutions (Windham, NH) was retained by KFx as an independent 3rd party to evaluate the test site and to prepare a test plan that would guide plant operations and KFx during testing. The objectives of the test were:

Satisfy plant concerns about handling, storing and feeding K-Fuel.
Show that K-Fuel could impact emissions performance, especially with respect to reducing mercury (Hg) and nitrogen oxides (NOx) emissions as compared to the Wyodak mine PRB coal currently burned at this site.
Quantify any operating benefits that K-Fuel may provide to plants currently burning PRB coals.

The Neil Simpson I boiler used for this test burn is designed to burn a relatively high moisture coal (~30%) produced at the nearby Wyodak mine. The boiler was built in 1969 and is nominally rated at 22 megawatts (MW) of electrical production, which is considered small by today's standards. The plant does not have the sophisticated burner or emissions control and monitoring systems of today's larger plants, which range in size up to 1300 MW. For purposes of controlling emissions, the plant does not have a sulfur dioxide (SO2) scrubbing unit but does have an electrostatic precipitator (ESP). The ESP unit is used to control particulate emissions. During operations, the minimum amount of excess air is used to ensure complete combustion. Flame and burner characteristics, among other factors, will determine how much excess air is required.

Air Pollution Testing Inc (APT) of Arvada, CO was retained to provide stack monitors for oxygen (O2), NOx, carbon monoxide (CO), and SO2 emissions. APT also sent a crew to perform Hg sampling using the Ontario Hydro Method recognized by the EPA. Boiler performance data were obtained by Quinapoxet from the Unit I control room.

Testing began during the first week of March 2006. As is standard practice in this type of test burn, the first part of the test involved slowly ramping up the percentage of K-Fuel in the coal feed blend so that operators could become comfortable with how the new fuel handles and feeds. After successfully operating on 100% K-Fuel™, the plant officially accepted the K-Fuel on March 7, 2006. At that point, the test was moved on to part 2: Baseline (Wyodak) coal tests.

The Neil Simpson I boiler normally uses a blended coal consisting of two seams from the Wyodak Mine (Anderson and Canyon). The blended product has a heating value of ~8000 Btu/lb and sulfur content of about 0.5 percent. The first several days of the Wyodak Baseline test were spent characterizing the NOx and CO emissions from the Wyodak coal and then sampling the Hg concentrations in the flue gas before the ESP to understand how these emissions are affected by the normal range of boiler operation at about 20 to 22 MW of power generation.

Results from the baseline coal tests showed that this boiler normally operates with relatively high CO emissions and NOx emissions of around 200 parts per million (ppm), when using the industry standard correction to 3% O2. The boiler airflow was increased to minimize the CO, but the underlying cause was poor flame stability and long flames in the lower burner row. When CO was reduced to below 500 ppm, NOx increased above 200 ppm.

Burning K-Fuel greatly improved the flame stability and shortened the flame length on the lower burners, while providing a brighter flame. The improvement can be attributed to K-Fuel's lower moisture content, which makes for a hotter flame. In fairness, the improper fuel-air mixing experienced by these burners may also be corrected through tuning, but the boiler controls did not allow operators to change burner settings. K-Fuel just made combustion control easier without operator intervention.

K-Fuel produced lower NOx emissions for this boiler by bringing the lower burner flames closer to the burners. Close ignition is required to assure that fuel-bound nitrogen is driven from the coal before the majority of the combustion air is available to convert it to NOx. The ignition instability described above allows early entrainment of combustion air and additional NOx formation. NOx was reduced by about 10 percent when burning K-Fuel at the same excess air as the baseline coal.

CO emissions were also reduced to acceptable levels when burning K-Fuel. Lower CO means that the boiler could be operated at a lower excess air percentage with K-Fuel. If NOx emissions are compared at equal CO emission levels (lower excess air), the NOx emissions were reduced by 15 to 22 percent with K-Fuel.

As stated before, the plant blends a higher sulfur coal seam with a lower sulfur coal seam to achieve sulfur dioxide emission compliance. Since the K-Fuel™ was produced from only the lower sulfur seam, the test burn showed a 40 percent reduction of stack emissions and the analysis of the coal from the feeder into the pulverizer.

Mercury emissions test results were somewhat ambiguous as there were problems (not related to the K-Fuel) with the testing equipment. As such, it is not possible to make definitive conclusions relative to mercury emissions. The mercury content of the Wyodak coal burned at Neil Simpson I during the baseline testing contained an average of 9.76 lb/trillion Btu's (TBtu) of mercury. The K-Fuel burned at Neil Simpson I had an average mercury content of 2.9 lb/TBtu (please see Appendix A for detailed fuel analysis). This means that the potential mercury emission is approximately 70% percent lower with K-Fuel. However, since some of the mercury can be absorbed on particles and become captured in the electrostatic precipitator, fuel-bound Hg content can only be used as a guideline. If the flue gas measurements are to be believed, then K-Fuel produced about 54 percent less mercury emissions than the baseline coal. Since some of the flue gas mercury samples had problems with contaminated glassware used in the testing analysis, Quinapoxet would recommend focusing on the fuel mercury results for now.

In all, over 1910 tons of K-Fuel were delivered and consumed at Neil Simpson Unit I. After this test burn, KFx can report the following conclusions to potential customers considering switching to K-Fuel:

The test results at Neil Simpson Unit I showed that K-Fuel could be delivered, conveyed, stored and fed to the mills just like any other commercial coal. In this case, K-Fuel used a truck delivery system, but fuel handling should be just as successful for plants that use unit trains or mine conveyors to deliver the fuel. Specifically, dust, size distribution, and grindability were all comparable to the Wyodak mine coal normally used at this plant. K-Fuel bulk density was slightly lower than the Wyodak coal, but that difference did not affect handling.
K-Fuel was able to partially overcome CO emission and flame shaping problems associated with current operating limitations at Neil Simpson Unit I. We do not expect such improvements at other plants with more sophisticated controls systems and the combustion system (mills and burners) are well tuned. However, it is clear that boilers that now burn PRB coals will have additional mill capacity when they burn K-Fuel. This will give operators much more flexibility to adjust burner air-fuel ratios for better flames and lower NOx emissions.
NOx emissions were reduced by 10 to 22 percent when burning K-Fuel at Neil Simpson I as long as the fuel heating value was above about 9500 Btu/lb. Less NOx reduction would be achieved with a well-tuned, low-NOx burner system of similar design. However, boilers with more burners and pulverizers can reduce NOx by biasing coal and air to certain burners to create larger regions of reducing (oxygen deficient) conditions. K-Fuel will help in that effort because it contains no iron and little sulfur to cause corrosion, and because it ignites easily to keep the flames steady. K-Fuel may also reduce NOx in boilers that use overfire air to control NOx (more stable flames may allow the use of more overfire air before reaching operating limits), but this benefit cannot be quantified without testing K-Fuel on such a boiler.
It is not practical for a mine mouth plant such as Neil Simpson to switch permanently to K-Fuel, but a similar plant now burning PRB coal would have to evaluate whether its mill controls can adjust to the higher Btu, lower moisture levels. Neil Simpson I operations chose not to make this modification for this test, but other test sites should be made aware of the issue so that they can decide how they want to deal with potential increases in mill outlet temperature in conjunction with changes in fuel throughput. Plants which burn a comparable coal in terms of moisture and Btu should not notice a significant difference.

In summary, K-Fuel proved to be a superior alternative to Wyodak mine coal at Neil Simpson. Results showed lower NOx emissions and significantly less Hg emissions. Flame stability improved and CO emissions also decreased with K-Fuel.

The only operating issue that bothered the operators and led to a potentially unsafe situation was predictable but unanticipated increase in the mill outlet temperatures because K-Fuel contains less moisture to evaporate during coal grinding. However, the temperature increase was within the range that the mill control system was designed to handle, and available tempering (cold) air should have been sufficient to maintain this temperature below 150 F where premature coal ignition is not a problem. In fact, mill 1 had no problem with controlling its outlet temperature.

With this issue in mind, K-Fuel is ready to step up to larger demonstrations on a wider variety of boiler and burner types. The next step should be a unit train sized test with a 100 to 150-MW boiler, preferably with overfire air, in order to show the maximum NOx-control potential of the product.

Notes:

Attachments highlight performance data on the K-Fuel process
In September 2005, KFx entered into an agreement with Black Hills Power for a test burn of K-Fuel at Black Hills Power's Neil Simpson station. Black Hills Power did not participate in the evaluation or documentation of the test burn results.
The author of this report was Mr. Steve Johnson, who is President of Quinapoxet Solutions and a valued Partner to the Power Generation Industry in the areas of coal combustion, NOx and Hg emission control. He has been developing, designing, and applying combustion and NOx-control technologies for over 30 years.

KFx PERSPECTIVE
For additional clarification of the test results, Quinapoxet Solutions has agreed to allow KFx to add some plant production information in this report.

The mercury content of the coal used for K-Fuel production contained an average of
7.9 lb/trillion Btus (TBtu) of mercury with the K-Fuel product exiting the processor at
2.6 lb/TBtu. The following tables provide the fuel analysis for these coals and product.

The Neil Simpson Unit I power plant normally burns a blend of two Wyodak coals, which is the baseline coal. When sampled from the feeder at the power plant, the baseline coal tested at 1.03 lbs of sulfur dioxide (SO2) per million Btus (MBtu). The KFx plant input was a single seam of coal which had a sulfur content of 0.72 lbs SO2/MBtu. K-Fuel produced at the KFx plant had a sulfur content of 0.64 lbs SO2/MBtu. This represents a 38% reduction in sulfur content from the baseline coal. The K-Fuel process accounted for 11% of the reduction in the sulfur content and the balance was due to the lower sulfur coal used as the feed stock for the K-Fuel process.

The difference between the 38% reduction stated above and the 40% SO2 reduction discussed in the Executive Summary is the difference between the feedstock testing and actual stack emissions.

The following fuel analysis was performed on the actual product from the K-Fuel™ Plant located at the Fort Union Mine in Gillette, Wyoming. The run period begins March 6 and ends March 18, 2006.

Run Purpose: Product for BHP Neil Simpson Station Test Burn

Coal Feed Origin: Wyodak Mine Canyon Seam Pit Run

Fuel Analysis: Feedstock Input to the K-Fuel™ Processor (see PDF document for breakdown)

Fuel Analysis: Product Exiting the K-Fuel™ Processor (see PDF document for breakdown)

Home

About Evergreen and K-Fuel®