Bryan M. Jenkins

The behavior of inorganic material in biomass-fired power boilers : field and laboratory experiences

This paper highlights some of the major findings of the Alkali Deposits Investigation, a collaborative effort to understand the causes of unmanageable ash deposits in biomass-fired electric power boilers. A group of interested industrial institutions and the US DOE Energy Efficiency and Renewable Energy Offices Biomass Power Program through the National Renewable Energy Laboratory jointly sponsored the project. The industries contributed both funding and, in most cases, use of facilities to the project and included Mendota Biomass Power and Woodland Biomass Power (both associated with Thermo Electron Energy Systems), CMS Generation Operating (formerly Hydra-Co Operations), Wheelabrator, Shasta and Hudson Energy, Sithe Energy, Delano Energy, the Electric Power Research Institute, Foster Wheeler Development, and Elkraft Power of Denmark. Research contracts with Thomas R. Miles Consulting Design Engineers, Sandia National Laboratories, and The National Renewable Energy Laboratories provided the government portion of the funding. In addition, the University of California at Davis and the Bureau of Mines performed significant work in close collaboration with the other researchers. This summary highlights the major findings of the project more thoroughly discussed in a recent report [2]. We highlight fuel properties, bench-scale combustion tests, a framework for considering ash deposition processes, pilot-scale tests of biomass fuels, and field tests in commercially operating biomass power generation stations. Detailed chemical analyses of 11 biomass fuels representing a broad cross-section of commercially available fuels reveal their properties that relate to ash deposition tendencies. The fuels fall into three broad categories: (1) straws and grasses (herbaceous materials); (2) pits, shells, hulls and other agricultural by-products of a generally ligneous nature; and (3) woods and recycle fuels of commercial interest. Woods and wood-derived products represent the most commonly used biomass fuels. Herbaceous fuels contain silicon and potassium as their principal ash-forming constituents. They are also commonly high in chlorine relative to other biomass fuels. These properties portend potentially severe ash deposition problems at high or moderate combustion temperatures. The primary sources of these problems are shown to be: (1) the reaction of alkali with silica to form alkali silicates that melt or soften at low temperatures (can be lower than 700°C, depending on composition), and (2) the reaction of alkali with sulfur to form alkali sulfates on combustor heat transfer surfaces. Alkali material plays a central role in both processes. The mobility of alkali material, defined as its ability to come in physical contact with other materials, is measured using chemical extractive techniques. Potassium is the dominant source of alkali in most biomass fuels. The analyses below indicate that essentially, all of the biologically occurring alkali, in particular potassium, has high mobility. The non-biologically occurring alkali is present as soil contaminants and additives to the fuels, such as clay fillers used in paper production. This non-biologically occurring alkali exhibits far lower mobility than the biological fraction. The relative amounts of biologically vs. non-biologically occurring material depend on fuel type and fuel handling. In the fuels investigated here, the dominant form of alkali was biologically occurring potassium. Some traditional indicators of deposit behavior, most notably ash fusion temperatures, poorly predict ash behavior compared with a more mechanistic interpretation of the data. Many of the agricultural by-products also contain high potassium concentrations with equally high potassium mobility. Some woods, on the other hand, contain far less ash overall, differing by as much as a factor of 40 from high-ash straws, for example. In addition, the ash-forming constituents contain greater amounts of calcium with

Boiler deposits from firing biomass fuels

Abstract Alkali in the ash of annual crop biomass fuels creates serious fouling and slagging in conventional boilers. Even with the use of sorbents and other additives, power plants can fire only limited amounts of these fuels in combination with wood. The National Renewable Energy Laboratory (NREL). U.S. Department of Energy (DOE), and the biomass power industry conducted eight full-scale firing tests and several laboratory experiments to study the nature and occurrence of deposits. The goal was to increase the quantities of these biofuels which can be used. This paper describes the results of the laboratory and power plant tests which included: tracking and analyzing fuels and deposits by various methods; recording operating conditions; and extensive laboratory testing. These analyses have advanced the understanding of the role of minerals in the combustion of biomass, and their occurrence in biofuels. Deposits occur as a result of the boiler design, fuel properties and boiler operation. The limited furnace volume and high flue gas exit temperatures of most biomass boilers promote slag or deposits from biofuels which contain significant amounts of alkali, sulfur or chlorine and silica. All annual growth, whether from urban tree trimmings, annual crops and residues or energy crops contains sufficient volatile alkali, 0.34 kg GJ − (0.8 lb MMBtu −1 ) or more, to melt in combustion or vaporize and condense on boiler tubes and refractory. Special boiler designs are required for annual crops, including grasses and straws. Addition of magnesium oxide and other additives may be necessary to inhibit alkali volatilization while burning these biofuels.

On the properties of washed straw

The removal of troublesome elements in biomass to reduce slagging and fouling in furnaces and other thermal conversion systems was tested by washing (leaching) the fuel with water. Rice straw and wheat straw were washed by various techniques and analyzed for composition and ash fusibility. Potassium, sodium, and chlorine were easily removed in both tap and distilled water. Total ash was reduced by about 10% in rice straw and up to 68% in wheat straw, although washing was more effective in increasing ash fusion temperatures for rice straw than for wheat straw due to the higher initial silica concentrations in rice straw. Untreated straw ash which fused below 1000°C was observed to become more refractory at higher temperatures when washed. Scanning electron microscopy of untreated and treated rice straw ashed at 1000°C revealed all untreated ash particles to be fused and glassy, while treated particles remained unfused, were heavily depleted in most elements other than Si, and displayed structures characteristic of original cellular morphology. The fusion temperatures of the straw ash were consistent with predicted temperatures from alkali oxide-silica phase systems based on the observed concentrations of elements in the ash. A simple attempt at simulating a possible full scale washing process was carried out by spraying the surface of a bed of straw with water for an arbitrary time of 1 min. This proved less effective in removing alkali metals and chlorine than soaking the samples in water, flushing water through them in a more controlled manner, or leaving the straw exposed in the field to natural precipitation. Electrical conductivity measurements of leachate revealed that extraction was mostly complete after application of 0.04 l g−1, equivalent to 24 mm of precipitation over uniformly spread rice straw. Full scale furnace experiments have not yet been conducted, and issues involving the practical application of the technique require further investigation, but these results suggest that fouling rates should decline for treated fuels compared to untreated fuels in conventional and advanced biomass power systems.

A comment on the optimal sizing of a biomass utilization facility under constant and variable cost scaling

Abstract Two solutions are obtained for the optimal size of a biomass utilization facility subject to an economy of scale in capital and non-fuel operating costs. The conventional assumption of constant-scaling parameter over all capacities leads to a larger optimum than if the scaling factor is increased asymptotically towards 1 with increasing capacity to reflect technical and economic constraints or risks. Lack of appropriate scaling data for larger size makes the question of general optimization for biomass facilities uncertain, and conclusions regarding development policies based on assumptions of constant cost scaling should be carefully tested.

Particle concentrations, gas-particle partitioning, and species intercorrelations for polycyclic aromatic hydrocarbons (PAH) emitted during biomass burning

Abstract Eight types of agricultural and forest fuels including 4 cereal crop residues and 4 wood fuels were burned in a combustion wind tunnel to simulate the open burning of biomass. Concentrations for 19 PAH species in particulate matter were found to range between 120 and 4000 mg kg−1, representing between 1 and 70% of total PAH emission. Weakly flaming spreading fires in the cereals were observed to produce higher levels of heavier PAH than more robust fires, with greater partitioning of PAH to the particle phase. Individual species concentrations appeared well correlated within groups based primarily on molecular weight, but no single species was observed to correlate with all others to serve as an indicator of PAH emission strength. Equilibrium gas-particle partitioning did not appear to be achieved within the 3–5 s residence time prior to sampling for sampling temperatures between 32 and 87°C, and in particular for the heavier species emitted from wood fuel pile fires with higher stack gas temperatures and shorter residence times. Total PAH emission, particle-phase concentrations, and fraction of PAH on particles were more strongly influenced by burning conditions than by fuel type.

Elemental characterization of particulate matter emitted from biomass burning: Wind tunnel derived source profiles for herbaceous and wood fuels

Particulate matter emitted from wind tunnel simulations of biomass burning for five herbaceous crop residues (rice, wheat and barley straws, corn stover, and sugar cane trash) and four wood fuels (walnut and almond prunings and ponderosa pine and Douglas fir slash) was collected and analyzed for major elements and water soluble species. Primary constituents of the particulate matter were C, K, Cl, and S. Carbon accounted for roughly 50% of the herbaceous fuel PM and about 70% for the wood fuels. For the herbaceous fuels, particulate matter from rice straw in the size range below 10 μm aerodynamic diameter (PM10) had the highest concentrations of both K (24%) and Cl, (17%) and barley straw PM10 contained the highest sulfur content (4%). K, Cl, and S were present in the PM of the wood fuels at reduced levels with maximum concentrations of 6.5% (almond prunings), 3% (walnut prunings), and 2% (almond prunings), respectively. Analysis of water soluble species indicated that ionic forms of K, Cl, and S made up the majority of these elements from all fuels. Element balances showed K, Cl, S, and N to have the highest recovery factors (fraction of fuel element found in the particulate matter) in the PM of the elements analyzed. In general, chlorine was the most efficiently recovered element for the herbaceous fuels (10 to 35%), whereas sulfur recovery was greatest for the wood fuels (25 to 45%). Unique potassium to elemental carbon ratios of 0.20 and 0.95 were computed for particulate matter (PM10 K/C(e)) from herbaceous and wood fuels, respectively. Similarly, in the size class below 2.5 μm, high-temperature elemental carbon to bromine (PM2.5 C(eht)/Br) ratios of ∼7.5, 43, and 150 were found for the herbaceous fuels, orchard prunings, and forest slash, respectively. The molar ratios of particulate phase bromine to gas phase CO 2 (PM10 Br/CO 2 ) are of the same order of magnitude as gas phase CH 3 Br/CO 2 reported by others.

Physical and Chemical Properties of Biomass Fuels

ABSTRACT HEATING value and fuel proximate analyses were determined for 62 kinds of biomass. Ultimate analyses were determined for 51 kinds of biomass. Biomass samples were selected from six categories: (1) field crop residues, (2) orchard prunings, (3) vineyard prunings, (4) food and fiber processing wastes, (5) forest residues, and (6) energy crops. Higher heating values ranged from 14.56 to 23.28 MJ/kg dry basis and were lowest for the field crop residues. Ash contents ranged from 0.17% to 24.36% dry weight basis. Nitrogen concentrations were found to be high in field crop residues, vineyard prunings, and in several types of food and fiber processing wastes. Correlation models were developed relating higher heating value to ash, volatiles, carbon, hydrogen, and oxygen.

Characteristics and biogas production potential of municipal solid wastes pretreated with a rotary drum reactor.

This study was conducted to determine the characteristics and biogas production potential of organic materials separated from municipal solid wastes using a rotary drum reactor (RDR) process. Four different types of wastes were first pretreated with a commercial RDR system at different retention times (1, 2 and 3 d) and the organic fractions were tested with batch anaerobic digesters with 2.6 g VS L(-1) initial loading. The four types of waste were: municipal solid waste (MSW), a mixture of MSW and paper waste, a mixture of MSW and biosolids, and a mixture of paper and biosolids. After 20 d of thermophilic digestion (50+/-1 degrees C), it was found that the biogas yields of the above materials were in the range of 457-557 mL g VS(-1) and the biogas contained 57.3-60.6% methane. The total solid and volatile solid reductions ranged from 50.2% to 65.0% and 51.8% to 66.8%, respectively. For each material, the change of retention time in the RDR from 1 to 3d did not show significant (alpha=0.05) influence on the biogas yields of the recovered organic materials. Further studies are needed to determine the minimum retention time requirements in the RDR system to achieve effective separation of organic from inorganic materials and produce suitable feedstock for anaerobic digesters.

Fuel gas enhancement by controlled landfilling of municipal solid waste

Abstract The controllable processing options available for maximum and accelerated fuel gas production from municipal solid waste (MSW) were studied in the laboratory in “controlled” landfill test cells. Fuel gas production was enhanced substantially compared to conventional “uncontrolled” sanitary landfills. Up to 0.142 m 3 CH 4 /dry kg MSW (4600 ft 3 CH 4 /dry ton) was obtained in just 3 months under the simulated landfill conditions, equivalent to a conversion of more than 50% of the biodegradable MSW fraction to fuel gas. Twenty-six controlled landfill experiments were performed in up to 200-litre (55 gallon) capacity test cells for periods of up to 410 days. Compaction to 712 kg/m 3 (1200 pounds/cubic yard) was used in all tests. Of the enhancement parameters investigated, the following significantly increased the fuel gas yield and production rate: (1) increasing the moisture content of the MSW; (2) controlling the pH with CaCO, buffer; (3) increasing the population of anaerobic microorganisms; (4) increasing the nutrients (mainly nitrogen and phosphorus); and (5) increasing overall distribution of the materials added throughout the MSW. Digested sewage effluent supplied the nutrients, the inoculum of microorganisms, and the moisture. Both static and leachate recirculation modes were also studied; recycle resulted in slightly higher fuel gas yields. Further, experiments were run at compactions from 593 to 949 kg/m 3 (1000 to 1600 pounds/cubic yard); percent MSW conversions to fuel gas were constant, but higher fuel gas production per unit volume was observed with increasing compaction.

Kinetic modeling for enzymatic hydrolysis of pretreated creeping wild ryegrass.

A semimechanistic multi‐reaction kinetic model was developed to describe the enzymatic hydrolysis of a lignocellulosic biomass, creeping wild ryegrass (CWR; Leymus triticoides). This model incorporated one homogeneous reaction of cellobiose‐to‐glucose and two heterogeneous reactions of cellulose‐to‐cellobiose and cellulose‐to‐glucose. Adsorption of cellulase onto pretreated CWR during enzymatic hydrolysis was modeled via a Langmuir adsorption isotherm. This is the first kinetic model which incorporated the negative role of lignin (nonproductive adsorption) using a Langmuir‐type isotherm adsorption of cellulase onto lignin. The model also reflected the competitive inhibitions of cellulase by glucose and cellobiose. The Matlab optimization function of “lsqnonlin” was used to fit the model and estimate kinetic parameters based on experimental data generated under typical conditions (8% solid loading and 15 FPU/g‐cellulose enzyme concentration without the addition of background sugars). The model showed high fidelity for predicting cellulose hydrolysis behavior over a broad range of solid loading (4–12%, w/w, dry basis), enzyme concentration (15–150 FPU/ g‐cellulose), sugar inhibition (glucose of 30 and 60 mg/mL and cellobiose of 10 mg/mL). In addition, sensitivity analysis showed that the incorporation of the nonproductive adsorption of cellulase onto lignin significantly improved the predictability of the kinetic model. Our model can serve as a robust tool for developing kinetic models for system optimization of enzymatic hydrolysis, hydrolysis reactor design, and/or other hydrolysis systems with different type of enzymes and substrates. Biotechnol. Bioeng. 2009;102: 1558–1569.