1. Scaling Assumptions

Stress testing your scaling assumptions with a model using defensible economic and biophysical values is critical to future business and project expansion. All figures are approximations for order-of-magnitude stress-testing, not quotes or engineering estimates.

Background¶

This note stress-tests the decision to scale a pilot community-scale sawmill — processing one MMBF/yr of small-diameter thinning material — into a 3-mill program, using the assumptions-based scaling framework in Salafsky et al. (2021) and outlined in Table 1. The central question is not “can one mill work?” but “what does the system actually require at 3 mills, and does the arithmetic hold?” The headline numbers: three MMBF/yr of lumber demands roughly 45,000 green tons/yr of small-diameter logs, which requires thinning 2,250 acres/yr; the program employs ~54 FTE (36 at the mills, 18 in field thinning crews) at a payroll of $3.15M/yr, and needs $4.5M of installed mill capital. Because small-diameter thinning material is volume-hungry and labor-intensive, feedstock and staff together cost roughly 3x the value of the commodity lumber the mills produce.

Models & Calculations¶

Figure 1 shows revenue vs. operating expenses over 7 years and expansion from one to three mills. The shaded gap indicates the difference between revenue and expenses over this time.

Two scaling assumptions fail the stress test in ways that mirror the wetland example Salafsky et al. (2021) provided. First, replication buys no capital economies of scale: three small mills carry the same $1.50 per board foot of installed capital as one, versus $0.67/BF for a single large industrial line. Second, the feedstock and staff base is far larger than intuition suggests: 15 green tons of logs are needed per MBF of lumber (vs. 3.8 tons/MBF for large-log industrial mills), and the payroll alone ($3.15M/yr) exceeds total lumber revenue ($1.05M/yr). The exercise therefore surfaces that the program only pencils out with value-added products (flooring/timbers), coproduct revenue (chips, residue, biochar), or a recurring subsidy on the order of $3M/yr — and that treatment costs vary so widely by region (California mechanical thin at $1,742/ac vs. $500—800/ac western averages) that the planning numbers below should be re-run on local costs before committing Delyser et al., 2025.

The purpose, as in the source framework, is a quick order-of-magnitude reality check rather than a precise financial model: it should reveal which assumptions would sink the program, where to spend effort refining estimates, and how to budget program management, ramp-up, and failure risk at scale Salafsky et al., 2021.

A single community mill does not pencil out from lumber alone: feedstock + staff exceed gross lumber revenue by roughly 3x (Table 2). Most mills close that gap with higher-value products, such as flooring, dimensional lumber, coproducts, log-sort premiums, or timber-sale revenue sharing Service, 2008.

Program-level parameters added when going from pilot to scale are shown in Table 3.

Table 4 shows the summary model outputs for 1 vs. 3 mills. The staffing model uses direct-employ field crews (Table C rows 9—11); the contract-thinning figure is the market-rate alternative.

Lessons¶

1. No capital economies of scale by replication. Building three small mills triples capital and keeps unit capital at ~$1.50/BF — versus ~$0.67/BF for a 300-MMBF industrial line Forestnet, 2025. Small-scale replication buys redundancy and community benefits, not capital efficiency (A balancing-loop “limit to growth” in Senge’s terms Salafsky et al., 2021)

2. Small-diameter feedstock is volume-hungry. ~15 tons of green material per MBF of lumber (vs. ~3.8—4 tons/MBF for large-log industrial operations Forestnet, 2025Service, 2008) means 3 MMBF of lumber demands ~45,000 tons/yr of thinning material and ~2,250 treated acres/yr — a large, sustained land base and a real constraint on scaling.

3. Staff dominates the cost curve. Payroll (~$3.15M/yr at full build) exceeds total lumber revenue (~$1.05M/yr) ~3-fold. The program cannot stand on commodity lumber; it requires coproduct revenue, value-added products, or ~$3M/yr of subsidy/co-finance — mirroring the conclusion that scaling required a large cash subsidy Salafsky et al., 2021.

4. Sensitivity levers (change any and re-run): lumber price ($250—500/MBF), value-added product mix (flooring/timbers sell near $1,000/MBF Service, 2008), feedstock cost ($25—45/ton), removal intensity per acre (10—40 tons/ac), and failure events (feedstock shortfalls, price drops) — all of which can swing the annual net by +/-$1M/yr. Regional treatment costs are the biggest wildcard: the western-US averages used here (~$500/ac, $21—28/ton) sit at the low end of the observed range, while California-specific costs (mechanical thin $1,742/ac, hand thin $1,088/ac, commercial-thin timber $59.43/GT Delyser et al., 2025) would roughly double-to-triple feedstock cost and wipe out any margin — so local figures must be re-run before committing Delyser et al., 2025.

5. Time frame. Even if approved, a 1-mill-every-3-yr ramp reaches full 3-MMBF capacity only in year 7 (mills at t=0, t+3, t+6) and requires sustained thinning throughout the ramp before feedstock stockpiles stabilize.