Grantee Research Project Results
Final Report: Procuring NHSM and Food processing by-products for Electricity Generation, Heating and Cooking applications
EPA Contract Number: 68HERC21C0028Title: Procuring NHSM and Food processing by-products for Electricity Generation, Heating and Cooking applications
Investigators: Jyamfi, Kwaku
Small Business: Farm to Flame Energy Inc.
EPA Contact: Richards, April
Phase: I
Project Period: March 1, 2021 through August 31, 2021
Project Amount: $99,999
RFA: Small Business Innovation Research (SBIR) - Phase I (2021) RFA Text | Recipients Lists
Research Category: Small Business Innovation Research (SBIR)
Description:
FTF Figure 1. Generator: Chamber Design, Combustion Testing, Blue flame present after testing
FTF leveraged its patented fuel processing and combustion technology to explore the viability of using NHSM and Food Processing By-products for Electricity Generation, Heating and Cooking Applications. The studied feedstocks were woody debris, yard trimmings and hay. Pre-treatment methods including chipping, drying, torrefying, and fine grinding were explored to increase the performance of the feedstocks as heating fuel. The resulting powdered fuel created using the pre-treatment methods was then burned through our patented combustion process designed for electricity generation, or pelletized and tested in a commercial stove. Combustion chamber geometry was designed using ANSYS fluent thermodynamic simulations, and built using Computer Numerical Control (CNC) machinery. Hands-on experimentation using pressurized air-streams throughout the combustion chamber was performed to optimize combustion. Electrical schematics, 3D modeling and computerized spatial configuration tools were used to design the integration of the patented combustion process into an electrical generator set-up. A series of custom-made metal pieces were designed and built to enable the coupling of the different components. The combustion chamber, steam boiler, alternator, recirculation tank, water treatment tank, electrical panel and ventilation was integrated into a 20’ container to study the portability of the generator. Data collected was used to refine pre-treatment parameters, in order to increase the fuel energy content, reducing the Carbon Monoxide (CO) and Volatile Organic Content (VOC) emissions created during combustion, reducing ash content and increasing pellet durability. This same data was used to understand the working conditions of the fuel processor, and ensure they complied with temperature, emissions and dust standards.
The energy return on investment (EROI) was calculated by adding the fuel processing and combustion equipment energy expenditure and dividing it against the electricity or heat output. Machine Learning (ML) models were developed to detect contamination in biomass waste streams, with the goal of screening future feedstock suppliers and having a flexible feedstock supply chain. Mechanical engineering analysis for the drying and torrefaction process and assessment of existing equipment suppliers was performed to build a cost model that would yield a cost vs benefit analysis.
Figure 2. Components of Lab-Scale Pellet Plant: Dryer Oven, Fine Grinder, Pellet Mill and Sensors
Summary/Accomplishments (Outputs/Outcomes):
Effective combustion in patented chamber required the combination of a specific pretreatment method, combustion chamber geometry and pressurized air-stream. This entailed reducing particle size below 100 micron, reducing feedstock moisture content below 5%, integrating primary, secondary and tertiary dilution holes and refining the air:fuel ratio to a 9:1 proportion. The presence of a blue flame at the core, and the lack of smoke, verified the ability of the combustion chamber design and pretreatment methods to achieve high thermal efficiency and complete combustion. Temperatures >800C in the commercial pellet stove were achieved with pellets made from softwood, yard trimmings and hay. The high temperature achieved during combustion in both the pellet stove and patented combustion process allowed for the decomposition of CO and other VOC’s into its non-toxic forms, mainly CO2 and H2O.
~1100% EROI was achieved through pelletization, and ~154% EROI through electrification, without torrefaction. Torrefaction provided pelletization benefits including increasing the energy content of the feedstock and reducing the moisture absorbed by the fuel during storage. However, it significantly decreased the EROI, and proved to require a mechanism, such as thermal oxidizers, to control the CO and VOC emitted. Fine grinding to micron-sized powder for pelletization proved to provide benefits including increased bulk density and durability. However, laboratory test results indicate bulk density surpassed the upper threshold (49.5 vs 48 lbs/ft3) and pellet durability surpassed necessary requirements (98 vs 95 PDI). Using yard trimmings and agricultural waste as fuel came with the added complication of increased ash content. Through 50/50 feedstock blending with wood debris, it was possible to create a yard trimming mixture that yielded an average ash content that complied with Pellet Fuels Institute standards (<6%). The presence of chlorine in pellet laboratory tests, due to run-off fertilizer in water, addresses the need for screening mechanisms of potentially toxic pollutants present in the biomass. Machine learning (ML) models developed for biomass quality assessment (QA) successfully detected >90% of glass, plastic and metal contaminants. Integrated model to detect unviable feedstocks was able to detect multiple instances of unviable feedstocks within a single image. Thermodynamic analysis for the rotary dryer yielded engineering parameters that will be used when sourcing equipment. Dynamic thermodynamic model showed a 15% energy discrepancy requirement between the most different climatic conditions, which will provide a basis for the requirements that our dryer will have to fulfill.
Pellet Figure 3.Testing: Wood Chipped, Fine Grinded and Torrefied pellets and Pellet Stove Combustion
Biomass electricity generator is fully built and scheduled to be piloted through October, 2021 in Selinsgrove, PA. Success will allow us to procure projects with 3 potential clients that have provided LOIs. Development of automated production plants with feedstock QA testing and tailored moisture reduction and particle size will be pursued to create generator fuel and pellets from mixed biomass waste materials. Need for mobile set up entices the placement of components of the fuel processor in shipping containers, so they can be modularly assembled. Abundant feedstock availability in landfills and pallet distribution centers enables FTF to get paid on a $/ton basis for feedstock removal, with relationships established with 3 landfills. After generator testing, client contracts to supply electricity, heat or pellets will be pursued to access necessary financing to purchase production equipment. Established relationships with equipment suppliers have enabled us to refine our cost-models in order to provide energy services, in the form of electricity and/or heat provision, to our clients on a PPA basis. FTF unlocked +$100,000 in additional funding from New Jersey Commission on Science Innovation and Technology, NYSERDA, and the Center of Excellence, in Syracuse, NY to continue developing our technology. Synergistic partnerships with university professors, investors, startup founders, and manufacturing specialists have been established for a successful Phase II project. Established relationships with funding organizations enable FTF to access necessary funding upon securing client contracts, and feedstock supply contracts.
Figure 4. Generator: Learning results, plastic, glass, and unviable fuel detection
Conclusions:
Satisfactory results set the stage for commercialization. Effectiveness of the patented combustion process resulted in its integration to a steam powered electricity generator. The ability to achieve high efficiency combustion while consuming a small amount of fuel (~60 lbs/hour) enables the development of affordable small electricity generation systems (30KW or bigger). Generator test results will enable FTF to explore the tradeoff between particle size and combustion efficiency to determine optimal particle size and chamber geometry. PFI compliant laboratory test results confirm our ability to produce high quality heating fuel from the biomass waste materials tested.
Although lab-scale equipment was used, which is not reflective of the increased efficiency achieved by commercial machinery, there was a favorable energy balance. Experimental findings showed it was unjustifiable to torrefy due to the high energy cost, the ability of non-torrefied fuel to reach the required 800C to reduce emissions, and the availability of affordable moisture-repellent storage options that eliminate the need for the hydrophobic properties created by torrefaction. Micron-sized fine grinding will be integrated into the electricity generation process, as it provides direct benefits related to increased thermal efficiency, emission reduction and preservation of equipment from corrosion or abrasion. However, the high energy cost and excessive bulk density of fine-grinding indicates fine-grinding is economically unjustifiable for pelletization. A component supplier database was built to assess potential fuel processing components, to effectively determine the optimal capacity for the entire fuel processor and set our supplier sourcing strategy accordingly. Refined cost-model indicates the viability of providing electricity on a power purchase agreement basis for <$0.15/kwh, and heat <$20 MMBTU. Future work is aimed at building a database that can be used to correlate the type of feedstock, pretreatment methods, and their combustion characteristics. Our goal is to build a predictive model that receives the available feedstocks to be used as heating fuel, as an input. Using regression coefficients, the model would output pretreatment methods that would enable our heating fuel to comply with specific parameters (e.g. moisture content, ash content, heating value) while minimizing the electricity consumed. Machine learning Image Detection models will be explored to provide useful information that can increase the predictive accuracy of the model and, thus optimize the pretreatment mechanisms.
ML Figure 5. Modeling: Learning results, plastic, glass, and unviable fuel detection
The perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.