Opportunities for Biomass Utilization in the Last Frontier

For the first time in more than 30 years, cofiring tests were performed at a utility-scale power plant in Alaska.
By Zackery Wright, Daisy Huang and David Nicholls | April 23, 2019

Cofiring coal and biomass is often a relatively simple process in which fuels are mixed prior to being combusted in a power plant. Although cofiring has been practiced for decades and is well-developed in Europe, it has been slow to catch on in North America. 

Coal is the primary fuel used to generate electricity in interior Alaska, including five small power plants, all grate systems less than 30 MW in size. total of about 600,000 tons of coal are combusted for power annually in Alaska, with local economic benefits associated with its mining, transporting and sales. The Usibelli mine, near Healy, Alaska, is the state’s only operating surface coal mine, producing about 1.5 million tons of coal per year and supplying all of the coal-burning facilities in the state. The coal power plants in interior Alaska serve a diverse client base, including a university, a municipality and two military bases. 

Fairbanks, Alaska, has great potential to utilize biomass cofiring, because the region’s power plants are all located near biomass resources.  Fairbanks’ power comes from mostly stoker-fired grate systems, which are the easiest to convert to cofiring. This is especially true when cofiring wood with coal at small percentages or with similar particle sizes; for example, wood chips and pea coal. Some of this resource is being used as a feedstock for a pellet mill in North Pole, Alaska, which is approximately 10 miles from downtown Fairbanks.  

The Aurora Power Plant, located in downtown Fairbanks, Alaska, was the site of the cofiring test burns. This facility has a 32-MW nameplate capacity, sells up to 25 MW to a local utility, and burns about 210,000 tons of coal per year. Net electrical generation is close to 180,000 megawatt-hours per year. In addition to electricity, the plant provides steam and hot water for a district heating system serving downtown Fairbanks. This network includes approximately 15 miles of buried pipeline and reaches approximately 50 buildings. 

Grate Systems Overview, Challenges
In grate-fired systems, the grate provides four main functions: to provide a platform for fuel drying, to burn the fuel, to distribute combustion air and dispose of ashes.  Grate systems, either stationary or traveling, are most common for small-scale energy production facilities (i.e., less than about 50 MW) that combust solid fuels. In grate systems, larger fuel particles of both coal and biomass can rest on the grate, while any fine particles that are included in the fuel mix will burn in suspension above the grate.

Research discussed in this article was completed at a combustion unit utilizing a conical fuel delivery system, which is a cone-shaped chamber located adjacent to the combustion chamber. The conical receives fuel until a certain predetermined level is reached, and then dumps fuel by gravity into the combustion chamber. An advantage of gravity-fed conicals is their reduced likelihood of jamming, a problem sometimes encountered with other types of fuel distribution systems.  However, a disadvantage of conicals is that the gravity feed can sometimes distribute different-sized particles unevenly in the combustion chamber. 

Wood Fuel Quality
Many fuel-related variables can influence the effectiveness of cofiring operations, such as moisture content, the percentage of biomass included in the fuel stream, relative energy content, fuel particle size distributions and quality (including the presence of bark, needles and dirt).  Fuel characteristics can also influence other operational aspects of cofiring, including fuel storage and conveying, mixing, distribution on the combustion grate, composition of combustion gasses and ash generation. In this study, the wood and fuel particles were approximately the same size, facilitating mixing and conveying.  The bulk density and particle geometry of the wood and coal can influence the mixing properties of the blend. In the combustion chamber, wood fuel can potentially burn more rapidly than the coal.  Thus, mill residues from an ongoing wood products facility are a preferred source for reliability and consistency, especially if available at low cost. Alaska is perhaps unique in that the local subbituminous coal has about the same Btu content as dried wood fuel (both are close to 8,000 Btu per lb.).  Therefore, mixing these fuels should require little adjustment to feed rates to achieve the same energy output.  This is not the case in the contiguous U.S., particularly where eastern coals are used, which may have as much as twice the energy content of biomass.

Wood/Coal Mixing, Conveying
Aurora Power Plant presently has limited outdoor storage area for wood fuel. This could be a consideration for other cofiring facilities (for example, municipal power plants) located centrally in or near a downtown area.  If Aurora were to cofire on a regular basis, or at higher percentages of biomass, additional storage and handling equipment would be needed. 

For small-scale grate systems, the easiest way to get a uniform fuel mixture is ususally to combine biomass and coal as it is transported inside the facility, often by conveyor. Despite the potential challenges of mixing wood and coal particles in real time during conveyance, our procedures resulted in very uniform fuel blends—credit to the power plant personnel in charge of fuel handling. It should be noted that although the procedure for introducing the biomass to the coal stream was simple for short tests, its manual nature makes it unsuitable for continuous cofiring, which would require a retrofit to the feed and mixing systems.

Goals of the study included: to test the proof of concept of cofiring quaking aspen wood chips and coal, under two different cofiring percentages; to measure the combustion gases in stack emissions during cofiring; and to identify any operational problems or challenges at the power plant associated with cofiring. 

Cofiring Procedures
Cofiring tests were conducted on two consecutive days in March 2015 at the Aurora Power Plant in downtown Fairbanks. Wood chips were provided by Superior Pellet Fuels in North Pole, fewer than 10 miles from the power plant. We utilized approximately 40 tons of clean, uniform aspen chips, sized to 1.5 inches maximum dimension or less. A front-end loader was used to transport the chips to the fuel feed system. Once fuel was inside the plant, a conveyer belt system was used to mix wood and coal uniformly to the desired ratio.

Cofiring tests were implemented at two levels:  2.4 percent by energy value for the low-level testing day, and 4.8 percent of energy value for the high-level testing. Baseline data representing the power plant’s standard combustion conditions (i.e., coal only) was collected for approximately one hour each morning.  Afterward, blended fuel (coal with biomass) was introduced at the prescribed rates. Actual cofiring rates were determined by collecting fuel samples at the accessible point closest to the combustion chamber, immediately before fuel was introduced into the conical. Wood was then separated from coal, and each were weighed. The coal energy content data was provided by plant personnel, and wood energy content was determined by oven-drying samples and testing them in a bomb calorimeter.

The composition of combustion gases were measured through the primary exhaust stack on the power plant roof. Combustion gas measurements were taken with a combination of three devices: a Bacharach brand PCA3 portable combustion gas analyzer (inserted into the stack), a TESTO brand analyzer (also inserted into the stack), and fixed, in-plant monitoring equipment.  The Bacharach analyzer was used to collect O2  (percentage), CO parts per million (ppm), CO with respect to O2  (ppm), combustion efficiency (percentage), CO2  (percentage), and excess air (percentage).  The TESTO unit was used to measure NO and NO2  (ppm) along with the same parameters as the Bacharach analyzer.  Opacity was measured by Aurora Power Plant’s in-house monitoring equipment. Data was collected either once every two minutes (Bacharach) or once every 30 seconds (TESTO). 

Low-level cofiring tests: Average flue gas concentrations for CO increased markedly (by 82.4 percent) versus coal-only combustion.  This was in contrast to CO2 concentrations, which decreased by 7.5 percent for low-level cofiring. Also for low-level cofiring, NO decreased by 20.32 ppm or 17.7 percent, and NO2  increased by 2.27ppm or 13.3 percent (all versus coal-only combustion). Other cofiring research (Tillman, et al. 2001) has indicated a greater than 1 percent reduction in NOx compounds for each 1 percent level of cofiring.  Finally, opacity increased slightly due to low-level cofiring, from an average of 5.7 to 5.9 percent (cofiring versus baseline). Thus, higher CO levels and particulate carryover can occur when cofiring wood and coal (versus burning coal only).

High-level cofiring tests: For high-level cofiring, average combustion gas CO content increased by 74.0 percent, from 127 to 222 parts per million (ppm), versus coal-only combustion. It is likely that higher CO levels were the result of either higher moisture content of wood (and the need to vaporize this moisture) or insufficient overfire air, which limited complete combustion. Similarly, CO2  concentrations were 21.6 percent higher. Also for high-level cofiring, NO increased slightly from 100.33ppm to 101.16ppm and NO2  decreased from 20.38 ppm to 18.42 ppm versus coal-only combustion.  Also for the high cofiring level, opacity increased slightly (from 6.0 to 6.1 percent) due to cofiring, however, it is unknown whether this is within the normal operating variation of the plant.

Several similarities between low- and high-level cofire tests were observed. First, flue gas CO increased substantially with the addition of biomass to the fuel mix.  Mean opacity increased, although only slightly, on both days when wood fuel was introduced, and 100 percent coal was no longer being burned.  Although these levels showed very little difference, opacity could become an important future consideration because of air quality concerns in Fairbanks (particularly during winter months). Thus, any cofiring strategies aimed to reduce opacity could become beneficial.

Power Plant Operations
Totential logistic and operational challenges when cofiring were also observed.  Our work in Alaska confirmed what numerous studies have observed—that cofiring biomass at low levels in grate combustion systems can be performed with relative ease, having only minor impacts on plant operations, including fuel storage, handling and performance. For low-level cofiring, the mixed fuel burned well with no issues or adjustments to combustion conditions. However, high-level cofiring was slightly more challenging, requiring careful control of feed rates to maintain uniform combustion across the grate. The wood’s high moisture content, estimated to be close to 40 percent green basis, was likely a factor influencing combustion during the high cofire tests.  In addition to generating greater incomplete combustion products such as CO, high moisture conditions also require more turbulent mixing in the flame zone for complete combustion to occur.  

Since the wood fuel was well-screened, individual particles were uniformly sized, and were only slightly larger than most coal particles. This could explain the relative ease of fuel mixing, and the uniform conditions occurring throughout conveying and combustion. The gravity-fed conical did a good job of distributing fuel uniformly across the grate, which was also likely related to the uniform chip size and moisture content.

The following specific comments were offered by the plant operator who monitored cofiring testing:

• Cofiring rates: In general, aspen chips cofired at the low rate introduced no problems or challenges. For the high cofiring rate, there was some minor segregation of aspen chips from coal.  Combustion was also somewhat more difficult due to greater level of moisture in the combustion chamber.

• Moisture content: lf chips had been at a lower moisture content, rather than fresh green, they might have burned more uniformly with more consistent steam load, even at the higher cofire rate.

• Spatial distribution of wood and coal: In general, the spatial distribution of wood and coal in the combustion chamber was quite uniform, and no problems were noted.

• Particle size range: If wood chips were oversized or too stringy, they could get caught in the under bunker part of the fuel conveying system, causing problems.  Chips or hog fuel up to a 2-inch size maximum dimension range could work well for the plant’s current fuel handling capability. Sawdust size should work well as long as it can be conveyed satisfactorily.

• Contaminants: Dirt, small rocks and other contaminants could be a problem, but not a cofiring showstopper. Any rocks in bottom ash would need to be run through a crusher, creating an additional processing step.

The tests at Aurora Power Plant demonstrated that cofiring relatively small amounts of biomass with coal is technically feasible with no equipment modifications needed, and minor impacts on plant operation.  An important consideration was that the cofire tests were done under ideal conditions using high-quality wood chips, purchased at higher prices than most run-of-the-mill-biomass that could be economically feasible on an ongoing basis.

Based on our research, we expect that cofiring low wood ratios would also be feasible at other four Fairbanks area power plants, since they are all of similar size, grate arrangements and coal types. Perhaps the most important finding was that CO levels increased on the order of 70 to 80 percent, versus coal-only combustion. This was likely due to the high moisture content of the aspen chips, resulting in less efficient combustion. However, this condition could potentially be mitigated as plant operators learn how to best fine-tune overfire air conditions. These factors point to the strong feasibility of cofiring green chips at levels up to about 10 percent of energy value, as was done in the study.

If cofiring were adopted on an ongoing basis, wood products facilities near Fairbanks would have a viable market for their waste wood residues, a condition likely to occur in many parts of the contiguous U.S. where power plants are located in forested regions.

Contact: David Nicholls
Forest Products Technologist, U.S. Forest Service