Feedstock Supply

 

 SUPERLONGTREE

Unlike conventional biofuel production systems, where feedstocks are produced from a defined area and tailored to a fixed conversion facility, beetle-killed trees are distributed irregularly over space and time, and new feedstock is created through episodic and largely unpredictable insect attacks.

Thus feedstock ‘production’ is not managed per se, but rather approached as a salvage operation. Accordingly, our objectives are to:

Locate and precisely inventory beetle-killed biomass

 

Field data collection

Beetle infestation in conifers is generally described in three phases based on visual and spectral characteristics: green, red and gray phase. To Identify and inventory feedstock from beetle-killed biomass, the effort will focus largely on infested conifer species that are in their red and gray phases. Red and gray phases can be detected using a variety of remote sensing approaches, including different airborne sensors, time-series analyses and derived vegetation/spectral indices. Extensive field data will need to be collected and is critical for training and testing the remote sensing methods. We will use fixed and variable radius plots to sample forest attributes and potential feedstock. Using these approaches and products, we will be able to detect and record beetle killed trees. Accurate inventory of feedstock will be critical for determining efficient and economically viable biofuel productions.

 

 

 

Quantify spatial and temporal distribution of feedstocks relative to ecosystem, topographic and infrastructure constraintsTeams_FeedstockSupply_pic2

Once beetle-killed trees have been located and inventoried, we will build off of these results to quantify the spatial and temporal distribution of feedstocks. By combining diameter-based allometric equations with integrated spatial data sets and correlative models, biomass estimates can be achieved at multiple scales. To accomplish this, we will employ a suite of tested and proven algorithms and spatial modeling techniques. These methods include classification and regression trees (CART), Boosted Regression Trees (BRT) and Maximum entropy (Maxent), among others. Using these models and integrating remote sensing, GIS layers, field observations and other spatial data sets, we will be able to generate spatial and temporal predictions of feedstocks. These results will be used to deliver a comprehensive digital atlas of current beetle-kill feedstocks for the region. These results will be combined with other geospatial data sets to support logistical analysis, conversion facility siting using decision support tools, and lifecycle and economic analyses.

Develop tools for rapid detection and quantification of newly occurring outbreaks that create new feedstock

After current distributions of beetle-killed feedstocks have been identified, we will develop procedures for forecasting new feedstocks and a method for rapid updating of feedstock inventory. To accomplish this, we will rely on the same spatial algorithms to develop temporal predictions. We will use historical data to calibrate methods and project model results 1-3 years out. Future projections will take into consideration host species distribution and condition, beetle dispersal limitations, and future environmental conditions. Under this framework, we will spatially and temporally model future beetle outbreaks to identity and quantify new feedstocks. These results will guide future logistics for efficient and economically viable biofuel productions.

 Teams_FeedstockSupply_pic3Teams_FeedstockSupply_pic4

Comments are closed