What FRED can and can't do

An honest catalog of the queries FRED is designed for, the ones it can partially address, and the questions to take elsewhere.

#What FRED handles well

The following are FRED's strongest capabilities:

  • Facility siting within California, anywhere with forest biomass coverage. FRED returns specific coordinates, supply curves, and cost projections.
  • Supply-radius analysis around a candidate location. Given a point and a demand, FRED computes the procurement radius needed to meet demand and the LCOE that results.
  • Cost-vs-fire tradeoff exploration. The Pareto framework is built for this. FRED can show the full curve, the leverage knee, and the cost premium at any alpha you specify.
  • Multi-year procurement planning with depletion modeling. FRED projects year-by-year supply and cost up to 15 years out, accounting for the fact that nearby clusters get harvested first and the radius must expand over time.
  • Regional summaries at the county, watershed (HUC8), or H3 hex level. Useful for screening multiple candidate regions before deep analysis.
  • Comparison queries between two or more candidate locations. FRED runs both analyses and presents a side-by-side summary.

#Partial coverage

FRED can answer the following, but with caveats worth knowing about:

  • Very long-horizon projections (15+ years). FRED's depletion model is deterministic and assumes prescriptions continue. Real-world procurement plans evolve. Treat 15+ year projections as rough trajectories, not commitments.
  • Custom prescriptions outside the C-BREC catalog. FRED uses the prescriptions already encoded in the cluster dataset. If you have a custom prescription, you can describe it, but the cost prediction may be less accurate.
  • Real-time pricing. FRED's cost model is from the harvest cost surrogate, which captures the engineering cost. It does not know about today's spot market for chip prices, fuel costs, or labor availability.
  • Permitting and regulatory feasibility. FRED can tell you the cost-optimal location; it cannot tell you whether you'll be able to get permits there. Use FRED to narrow the search, then apply local knowledge.
  • Specific landowner contracts. The cluster supply is on a mix of ownership types. FRED knows the rough split but cannot model individual landowner contracts.

#Out of scope

FRED currently does not handle the following:

  • States other than California. The cluster dataset, fire raster, and transport model are all California- specific. Other states require running the data prep pipeline against equivalent regional data.
  • Agricultural residues. FRED's supply is forest biomass only. Orchard residues, dairy waste, and crop residues are not in the cluster catalog.
  • Carbon accounting. FRED tracks BDT supply and cost. It does not produce carbon offset estimates, RPS compliance calculations, or life-cycle emissions analyses.
  • Detailed facility engineering. FRED tells you the biomass economics. It does not size the boiler, layout the yard, or design the chipping line.
  • Endangered species, habitat, or wilderness constraints. FRED's cluster polygons reflect the C-BREC inventory, which incorporates broad land use categories. Site-specific environmental review is a separate workflow.
CAUTION
FRED is decision support, not a decision system. The recommendations are grounded in published research and validated models, but they are inputs to your process. Real procurement decisions involve permitting, contracts, community engagement, and operational considerations FRED does not see.

#On the roadmap

Capabilities under active development:

  • Coverage expansion to Oregon and Washington using equivalent C-BREC-style inventories.
  • Custom prescription support for users with their own treatment specifications.
  • Sensitivity and scenario analysis built into the chat experience, so you can ask "what if diesel prices rise 20 percent" and FRED applies it.
  • Persistent project workspaces so analysis on one facility can be saved and revisited weeks later with comparisons to the original recommendation.

If you have a use case FRED almost handles but doesn't quite, email contact@biofred.us. The roadmap is driven by what users are actually trying to do.