Your Tax Equity Partner Just Asked: "How Confident Are You in This COD Date?"

Community solar developers and VPP aggregators lose financing when they can't defend their construction schedules. We translate equipment lead times and interconnection queue uncertainty into board-defensible diligence packs, before tax equity walks or lenders reprice your risk.

  • Submit your BOM + construction schedule (P6 XER export or CSV template)
  • Receive Schedule Realism Pack in 72 hours: procurement register, COD sensitivity, assumption ledger, risk register
  • Use outputs in tax equity diligence, lender presentations, or subscriber commitment decisions
See the diligence packs Request demo access
$15k-$25k typical tax equity diligence pack for 50 MW community solar project
72 hours turnaround from BOM + schedule submission to Schedule Realism Pack delivery
3-month COD slip from 12-week transformer delay can breach tax equity covenant (Dec 31 deadline)
128 weeks average transformer lead time (up from 52 weeks in 2019); DER inverters: 16-36 weeks
What Epsilon is
  • A diligence layer on top of Primavera P6 and MS Project
  • Audits procurement assumptions buried in your schedule
  • Produces artifacts schedules do not generate, such as registers, sensitivity, and ledgers
What Epsilon is not
  • A schedule authoring tool: we ingest XER exports and do not replace P6
  • A field execution platform (Procore, etc.)
  • A live project tracking dashboard
The Financing Risk Gap

Developers & Owners Are Flying Blind

Tax equity partners and lenders require COD risk analysis before close. But your BOM, schedule, and financial model don't talk to each other. By the time delays surface, it's too late to save the financing.

Tax Equity Won't Close Without COD Proof

50 MW community solar, $75M CAPEX, Dec 31 COD covenant. Transformer delay (+12 weeks) → 3-month slip → missed deadline → $45M tax equity tranche delayed. Schedule Realism Pack quantifies this exposure before close.

Lenders Reprice Risk You Can't See

100 MW VPP portfolio, $120M construction loan. Battery delay (+8 weeks) → Q2 → Q3 COD → miss summer peak → capacity payments drop 15% → DSCR 1.25x → 1.18x → covenant breach. Lenders demand $2M higher reserve account.

Subscriber Churn Is a Financing Risk

10 MW community solar, 600 subscribers, $240k acquisition cost. Queue delay (12 months) + transformer delay (8 weeks) → 20-month COD slip → 25% churn (150 subscribers) → $1.2M LTV drop. Should you pause subscriber acquisition?

Equipment Lead Times Are Not Static

Distribution transformers: 30-60 weeks (DER-scale). Transmission transformers: 100-144 weeks (utility-scale). Residential inverters: 16-28 weeks. Central inverters: 24-36 weeks. Your schedule needs provenance for every assumption.

Who it's for

Three Diligence Packs. Three Critical Decision Points.

We deliver Schedule Realism Packs at three financing trigger points: tax equity close, construction loan underwriting, and subscriber commitment decisions.

Community Solar Developers

DER Diligence Pack for Tax Equity

Defend your COD date to tax equity partners. 50 MW portfolio, $75M CAPEX, partnership flip structure. Transformer delay (+12 weeks) → 3-month slip → Dec 31 covenant breach. Schedule Realism Pack shows exposure + mitigations before close.

  • Procurement register (long-lead items, needed-by dates)
  • COD slip sensitivity ("tornado chart" ranking top 10 schedule killers)
  • Assumption ledger (provenance: quote / vendor / category default)
  • Risk register (exposure + mitigations: split procurement, expediting)
VPP Aggregators

VPP Revenue-at-Risk Analysis

Quantify how equipment delays impact VPP revenue. 100 MW portfolio, $120M construction loan. Battery delay (+8 weeks) → Q2 to Q3 COD → miss summer peak → PJM Capacity Performance non-delivery penalties trigger → DSCR breach. Get revenue-at-risk breakdown.

  • Queue + equipment delay scenarios (distribution-level POI)
  • VPP dispatch forecast (capacity + energy + ancillary revenue)
  • Subscriber economics (churn scenarios: 10%/20%/30%)
  • Feeder hosting capacity (voltage/thermal violation risk)
Project Controls / Schedulers

Schedule Realism Pack

Defend schedule assumptions to investment committees. Your P6 schedule says transformer arrives Month 10. We audit: "Category default 72 weeks (Wood Mackenzie); vendor quote 80 weeks valid thru Jun 2026." Get assumption provenance.

  • BOM ↔ Schedule binding map (which equipment gates which tasks)
  • Lead-time provenance (quote / vendor email / category default / override)
  • Delta narrative ("what changed since last update?")
  • P6 XER export compatibility (import schedule, export diligence layer)
How it works

Schedule Realism Pack. Delivered in 72 Hours.

Submit your BOM + construction schedule. We bind equipment lead times to your critical path, run slip scenarios, and deliver board-defensible outputs. Not a schedule tool, but a diligence layer that audits your P6 assumptions.

Step 1

Submit BOM + Schedule

Upload P6 XER export (or CSV template). Provide equipment BOM (from quotes/ERP). Include any vendor quotes or lead-time assumptions.

Step 2

BOM-Schedule Binding

We bind equipment lead times to procurement/construction tasks. Transformer (72 weeks) gates "Install Transformer" (Month 10). Audit provenance.

Step 3

Slip Scenarios + Sensitivity

Run what-if: Transformer +12 weeks → COD slips 3 months. Battery +8 weeks → Q2 → Q3 miss. Rank "top 10 schedule killers."

Step 4

Deliver Diligence Pack

PDF + CSV exports: Procurement register, COD sensitivity, assumption ledger, risk register, delta narrative. Use in tax equity / lender presentations.

Not a Schedule Tool. A Diligence Layer.

We don't replace Primavera P6 or MS Project. We ingest your XER export, audit procurement assumptions, and produce artifacts schedules don't generate: provenance, sensitivity, mitigations.

Assumption Provenance Tracking

Every lead time has a source: vendor quote (SMA Americas, valid thru Jun 2026), category default (Wood Mackenzie, Jan 2026), or manual override. Tax equity partners require this audit trail.

DER-Specific Equipment Catalog

Distribution transformers (5-15 MVA, 30-60 weeks). Residential inverters (Enphase, SolarEdge, 16-28 weeks). Utility batteries (Megapack, Fluence, 36-60 weeks). Community solar racking (16-24 weeks).

72-Hour Turnaround

Submit Monday morning. Receive Schedule Realism Pack Thursday afternoon. Pricing available upon request after initial scope review.

DER-at-Scale Segment

Community Solar + VPP: Different Equipment, Same Diligence Need

DER-at-scale projects (10-100 MW) face different supply chain risks than data centers (250+ MW). Residential inverters (16-28 weeks) vs. transmission transformers (128+ weeks). Distribution queue delays (1-3 years) vs. transmission queue (3-5 years). Same diligence framework, different equipment catalog.

Distribution-Level Equipment

Pad-mount transformers (5-15 MVA, 30-60 weeks). Residential inverters (Enphase IQ8, SolarEdge HD-Wave, 16-28 weeks). Community solar racking (Array Tech, Nextracker, 16-24 weeks).

Interconnection Queue Risk

Distribution POI delays (1-3 years). System impact study timelines and up-queue withdrawal cost allocation. Queue position #47 can trigger retroactive network upgrade costs if upstream projects drop out.

Subscriber Economics Impact

600 subscribers @ $85/month credit. 12-month delay → $240k lost revenue Year 1 + 25% churn (150 subscribers) → LTV drop $1.2M. Should you pause $240k acquisition spend until queue clears?

VPP Revenue-at-Risk

Battery delay → Q2 → Q3 COD → miss summer peak → PJM capacity payments drop 15% ($150k/MW-year → $127.5k). DSCR 1.25x → 1.18x → covenant breach. Lender requires $2M higher reserve account.

What You Receive

Six Core Deliverables. Board-Defensible Outputs.

Every Schedule Realism Pack includes these six outputs. Designed for tax equity diligence, lender presentations, and investment committee reviews.

Procurement Register Export

CSV + XLSX

Long-lead items filtered by threshold (default: 20+ weeks). Columns: SKU, category, qty, lead_time_weeks, needed_by_date, status. Sorted by needed_by_date ascending.

COD Slip Sensitivity Report

Tornado Chart

For each long-lead SKU, run slip what-if (+4, +8, +12 weeks) → measure COD impact. Rank "top 10 schedule killers" by max COD slip (weeks) descending.

Assumption Ledger (Provenance)

Audit Trail

Every lead time has a source: vendor quote (SMA Americas, valid thru Jun 2026), category default (Wood Mackenzie, Jan 2026), manual override. CSV export with date_recorded + confidence.

Risk Register Export

Exposure + Mitigations

Risk: Transformer delay. Exposure: 3-month COD slip. Mitigations: (1) Split order across 2 OEMs, (2) Expedite fee $85k, (3) Partial energization (Phase 1: 30 MW by Dec 31).

Delta Narrative Generator

Change Summary

Automated "what changed since last update?" Sections: New BOM lines, removed lines, changed lead times, COD forecast update (baseline/downside scenarios). Markdown + PDF.

BOM-Schedule Binding Map

CPM + Procurement

Which equipment gates which tasks. Transformer (72 weeks) → "Install Transformer" (Month 10). Inverter (24 weeks) → "Energization" (Month 12). Coverage: % of critical path tasks bound to procurement.

What the models show

Cross-scale findings from running the platform.

These findings emerge from analyzing real project data across the procurement-to-construction lifecycle. Each one exposes a blind spot that spreadsheets and siloed systems miss.

Equipment bundling changes the economics.

Procurement Strategy

A project that looks marginal on transformer pricing alone can be profitable when panel, inverter, and racking bundles are optimized together. Siloed procurement understates value.

Procurement and scheduling are not separable.

Operations

Equipment ordered without schedule awareness often arrives before the site is ready or after the critical path has shifted. Stopping at one stage misleads.

A BOM that looks good on paper can fail at the site.

Project Delivery

Specification match, delivery timing, and installer readiness are three different failure modes. A perfect BOM should be discounted by all three.

Supply chain shocks compound; they do not add.

Risk Analysis

Tariffs + port congestion + raw material shortages is a tail problem. Projects sized for a single P50 stressor often miss the P90 of the combination.

The last 10% of on-time delivery costs more than the first 90%.

Lead Time

Bringing every SKU in on schedule requires buffer inventory and expedited freight that most models ignore: what true project cost accounts for.

Supplier gaming emerges without being programmed.

Market Dynamics

Over multi-year relationships, quoted lead times can drift upward through payoff learning: a stress test for "trust but verify" procurement assumptions.

Substituting upstream does not fix a downstream constraint.

Logistics

Two suppliers can both be "qualified" and still fail to deliver if one ships to the wrong warehouse. Physical reality beats aggregate availability.

Optimistic lead-time assumptions oversell feasibility by 40-80%.

Planning

The same project can book at 18 months under base accounting but 28+ months under realistic lead times. Treating quoted dates as certain is a planning fiction.

BOMs without schedule visibility are unsafe at scale.

Execution

Blind procurement can drive overstock, stockouts, and cash flow violations. Schedule-level visibility is a requirement, not polish.

Together: project delivery needs cross-scale coherence, and every scale has a blind spot the next scale can see.

Advanced

Legacy Energy Modeling Capabilities

Project Epsilon also includes advanced energy system modeling engines originally developed for academic and research use. These remain available for power users and remain part of the technical foundation.

DC-PPA Optimizer

PyPSA LP

Optimal solar, wind, battery sizing for 24/7 CFE targets with hourly dispatch simulation and LCOE/CostStack analysis.

ISO-RTO Stress Testing

Monte Carlo

Dispatch simulation with EV growth, dunkelflaute, and forced outage scenarios. P50/P90 confidence bands for grid reliability.

Microgrid Financials

PyPSA + Finance

Site-level dispatch with PPA, VPP, and export revenue stacking. NPV, IRR, payback, and sensitivity analysis.

VPP Market ABM

Agent-Based

Multi-agent market simulation for virtual power plant bidding, reserves, and redispatch with strategic gaming detection.

Network Feeder Tools

DC/AC Power Flow

Radial line constraints, N-1 contingency analysis, and IEEE 33-bus AC power flow for hosting capacity studies.

Roadmap

Building the EPC Operating System

From BOM generator to marketplace. Our roadmap is designed to compound value at every phase.

BOM + CPM Core

Now 100+ SKUs, XER import/export, lead-time binding
Q2 2026 AI-assisted BOM generation from project specs
Foundation: the single source of truth for equipment and schedule

Live Pricing & Alerts

Now BloombergNEF / NREL SAM baseline indices
Q2 2026 BYOK distributor APIs + delay alert engine
Real-time visibility into what your BOM actually costs today

Finance Integration

Now CAPEX true-up with LCOE/IRR/NPV sensitivity
Q3 2026 Full project finance model with carrying costs
Connect procurement reality to lender conversations

Demand Aggregation

Q3 2026 $500M+ project pipeline visibility for suppliers
Q4 2026 Two-sided marketplace launch
Network effects: more projects attract more suppliers

Compliance & Reporting

Q3 2026 UFLPA and IRA domestic content tracking
Q4 2026 Automated compliance documentation
Regulatory risk is procurement risk

Enterprise Scale

Q4 2026 SSO, multi-tenant, custom integrations
2027 API platform for ERP and PMIS integration
Become the procurement layer, not just a tool
Climate risk + data

Acute hazards, chronic drift, and scenario uncertainty.

Today, many modules use synthetic weather and static priors. The roadmap maps external datasets to engines: a fidelity lift, not a rewrite of the scenario contract.

Risk or signal Primary data sources Modules What it changes
Solar resource variability NREL NSRDB, NASA POWER, PVGIS DC-PPA · Microgrid · ISO Hourly 8760s, multi-year spread vs synthetic curves
Wind resource variability ERA5, NREL WIND DC-PPA · ISO Correlated shortfalls, multi-site wind
Extreme heat & cooling NOAA LCD, CMIP6 downscaled Microgrid · DC-PPA · ISO PUE and thermal derates under heat
Grid carbon intensity WattTime, eGRID, Elec. Maps DC-PPA · Microgrid Marginal emissions vs static kg/MWh

Resource loaders

Locational keys into profile-style paths replace pure synthetic curves where datasets attach.

Outage priors

ISO-RTO Monte Carlo can take hazard-weighted derates (e.g. wildfire) instead of flat rates.

Scenario tags

Optional climate_scenario in handoff keeps load and resource pairs aligned across stress runs.