Photometrics AI Financials

Purpose: Single source of truth for all financial calculations, quantitative claims, and their derivations.

Core principle: Never state a number as fact unless it has verified logic and sources documented here.

Quick Reference: Key Numbers

Status Key:

• •✅ VERIFIED — Backed by cited source
• •🔶 DERIVED — Calculated from verified inputs
• •⚠️ ESTIMATE — Interpolated or approximated
• •🔴 PLACEHOLDER — Needs source; do not cite as fact

Unknown ≠ Zero: Defending Value Estimates

The Principle

The absence of a precise number does not mean the value is $0. It means the value is somewhere between $0 and some upper bound. An estimate rooted in published studies and documented methodology is almost certainly closer to the real number than zero.

Someone who dismisses a value category is not being conservative. They are implicitly claiming the value is $0 — and that claim requires the same justification as any other.

Why This Matters

The financial model above contains values at every level of the status system. Skeptical stakeholders — PEs, utility program managers, municipal finance directors — sometimes dismiss categories they can't directly measure or capture. The implicit logic: "I can't quantify this precisely, so I'll treat it as zero."

This is analytically wrong. Consider the components of the $51.09/light/year municipal value:

Each of these dismissals contradicts published research. The question is never "Is this value exactly $10.81?" — it's "Is this value closer to $10.81 or to $0?" When the estimate is built from BJS expenditure data, LAPD crime records, and peer-reviewed methodology, the answer is clear.

How This Connects to the Status System

The status system isn't a weakness to apologize for. DERIVED and ESTIMATE values are honest about their methodology — which makes them more defensible than unstated assumptions of zero.

The Stakeholder Capture Problem

A common dismissal: "The municipality can't capture that savings directly." This confuses value capture with value existence.

Example: A city with unmetered street lighting can't see energy savings on a utility bill. A skeptic concludes the savings don't exist. But the energy reduction is physical — fewer kilowatt-hours consumed, lower thermal stress on components, reduced grid load. The value accrues somewhere: the utility sees lower demand, the grid sees reduced congestion, the luminaires last longer, the atmosphere absorbs less carbon. Inability to invoice for a benefit is not evidence the benefit is zero.

When a stakeholder says "we can't capture that," the correct response is to identify who does capture it and whether it still supports the business case. Value that accrues to a different stakeholder is still value — and often still motivates the decision.

When to Invoke This Principle

• •A stakeholder dismisses quality-of-life values (crime, traffic safety) as "soft"
• •Someone treats DERIVED or ESTIMATE values as equivalent to PLACEHOLDER
• •A PE or engineer rejects ROI because one component can't be directly metered
• •A reviewer says "you can't prove the exact number" as grounds for using zero
• •Anyone frames the choice as "perfectly measured or worthless"

Response Framing

When challenged, the structure is:

1. •Acknowledge the uncertainty. "You're right that $10.81 is an estimate, not a measurement."
2. •Reject the false alternative. "But the alternative isn't a better number — it's an implicit claim that the value is $0."
3. •Show the methodology. "Our estimate uses [source], [methodology], [conservative assumptions]. Here's the derivation."
4. •Shift the burden. "If you believe the value is closer to $0 than to $10.81, what's the basis for that?"
5. •Offer the range. "Even if you halve every estimate in the model, the ROI is still under 12 months."

Operating Hours: Two Numbers for Two Purposes

Street lights operate dusk-to-dawn, but different calculations require different hour values:

Why the difference?

• •4,100 hours is the regulatory value embedded in utility tariffs. When calculating dollar savings from reduced consumption, use this number because it matches how utilities bill.
• •4,165 hours is the astronomical reality of dusk-to-dawn operation. When calculating actual energy consumed or CO2 avoided, use this number because it reflects physics.

In practice:

• •Energy cost savings: Use 4,100 hrs → 205 kWh/light/year
• •National energy impact / CO2: Use 4,165 hrs → 208 kWh/light/year

Reference Files

⚠️ CRITICAL: Always Check Numbers Audit First

Before citing ANY number, read Numbers Audit.

Each number shows: the claim, derivation logic, source citations, and verification status.

• •Numbers Audit — Derivation logic and sources for ALL quantitative claims
• •Financial Model — Full methodology for $51.09 municipal + $10.18 utility calculations

Source Documents

Source PDFs and images are in references/sources/:

Value Breakdown: Municipal ($51.09/light/year)

Asset Management: $30.44

Quality of Life: $20.65

DSM Value Attribution: Can be categorized as municipal benefit (city receives payments), utility benefit (grid flexibility), or quality of life benefit (rate stability, blackout reduction).

Value Breakdown: Utility ($10.18/light/year)

Based on CPUC 2024 ACC Electric Model v1b, Climate Zone 10 (Riverside, CA).

Key Finding: Nighttime hours have zero capacity value (Gen Capacity, Transmission, Distribution = $0/MWh). Overnight avoided costs are GHG + Energy only (~$130-150/MWh).

Dimming Schedule (35% savings):

Calculation: Hour-by-hour ACC rates × hours × dimming factor, using CZ10 twilight data.

• •Fleet: 20,000 lights × 50W = 1 MW
• •Annual ACC benefit: $203,699
• •Per light: $10.18/light/year

Sources: 2024-ACC-Electric-Model-v1b.xlsb, AccStreetLightingAnalysis.xlsx, cz10_riverside_2026.csv

Note: California-specific. Other jurisdictions require different methodology.

Pricing & ROI

SaaS Pricing: $3-12/light/year

Based on:

• •Deployment size (volume discounts)
• •Service level (self-service vs managed)
• •Feature set (basic vs full dynamic scheduling)

ROI Calculation

Value delivered: $51.09/light/year
Cost: $3-12/light/year
Net benefit: $39.09-48.09/light/year
ROI: <12 months

Example: 50,000-light city

• •Annual value: 50,000 × $51.09 = $2.55M
• •Annual cost: 50,000 × $6/light = $300K (mid-range)
• •Net savings: $2.25M/year

Pilot Structure

Market Size

Global Streetlights

US Streetlights

California Streetlights

Baseline: AB 719 / CPUC Data (2013)

The most defensible single source for IOU streetlight counts is the AB 719 bill analysis, which cites data provided by the CPUC to the California Legislature:

Source: AB 719 Assembly Bill Analysis (April 2013) — data provided by the CPUC.

• •File: references/sources/AB 719 Assembly Bill - Bill Analysis.pdf

Cross-References

SCE note: The Edison article states SCE owns ~450,000; cities own a comparable number in SCE territory. This implies ~900,000 total in SCE territory today — above the 2013 CPUC baseline of 768,669, suggesting growth and/or broader counting methodology.

PG&E note: PG&E's website states "nearly 670,000" total in their territory, below the 2013 CPUC figure of 729,585. This likely reflects city acquisitions of utility-owned lights (per AB 719 programs) removing lights from PG&E's tracked inventory, not a reduction in total lights.

Estimation Methodology: Population-Proportional Inflation

To estimate current streetlight counts from the 2013 CPUC baseline, a GIS analysis was performed (January 2026) using territory-specific population growth factors.

Method:

1. •Spatial join of 9,129 California census tract centroids to CEC utility service territory polygons
2. •Population assigned per territory using ACS 2023 5-year tract-level data
3. •2013 population estimated by deflating 2023 tract populations using county-level ACS 2013/2023 growth ratios (counties have stable boundaries across census vintages)
4. •AB 719 streetlight counts inflated by each territory's population growth factor
5. •Non-IOU territory lights estimated using the IOU-derived lights-per-capita ratio

Data Sources:

• •Utility territories: CEC ElectricLoadServingEntities_IOU_POU (ArcGIS FeatureServer, 50 retail utilities after removing 3 non-retail overlays)
• •Census tracts: Census TIGERweb ACS2023 boundaries (9,129 tracts)
• •Population: ACS 2023 5-year (tract level), ACS 2013 5-year (county level for growth factors)
• •Streetlight baseline: AB 719 Bill Analysis (April 2013), CPUC data — File: references/sources/AB 719 Assembly Bill - Bill Analysis.pdf

Data cleaning: Three non-retail overlay entities were removed to prevent double-counting: Power and Water Resource Pooling Authority (wholesale, overlaps PG&E), Metropolitan Water District of So. Cal (water district, overlaps LADWP), Eastside Power Authority (JPA, overlaps PG&E). After removal, 255 tracts still had dual matches (San Francisco tracts in both PG&E and Hetch Hetchy territory); these were assigned to the first match.

Reproducibility: Analysis script at references/sources/gis_analysis/run_analysis.py, results at references/sources/gis_analysis/streetlight_estimation_results.json.

GIS Analysis Results: Population by Territory

California population: 39.24M (ACS 2023). Previous figure of 52.2M was incorrect.

Current Streetlight Estimates (🔶 DERIVED)

Lights-per-capita ratio (IOU): 0.0566 (1 light per 17.7 people)

Non-IOU estimation method: Applied the IOU lights-per-capita ratio to non-IOU territory populations. This assumes similar streetlight density — a simplification, since urban munis (LADWP) likely have higher density and rural districts lower.

Top non-IOU utilities (estimated lights):

Validation: LADWP estimate of ~216K matches the known 223K from the LA Lights Strategic Plan 2020-2025, supporting the lights-per-capita ratio.

Known Limitations

• •Population ≠ streetlights: The assumption that streetlight density scales linearly with population is a simplification. Urban cores have more lights per capita; rural areas fewer. Highway lighting adds lights without corresponding residential population.
• •2013 baseline age: The AB 719 data is from 2013. Since then, cities have acquired utility-owned lights (AB 719 programs), new developments have added lights, and some areas may have consolidated. Population growth is a proxy for net new infrastructure.
• •SCE gap: The SCE estimate (~792K) is below the Edison article's implied ~900K total. This may reflect SCE territory having above-average lighting density, or the Edison figure capturing lights not in the 2013 CPUC count.
• •Non-IOU assumption: Applying the IOU ratio to munis is approximate. Where known data exists (LADWP 223K), it should override the estimate.

National Energy Impact

Claim: 4,372 GWh saved annually (408,000 homes equivalent)

Calculation:

Annual kWh per light: 50W × 11.4 hrs × 365.25 = 208.19 kWh
US total (60M lights): 208.19 × 60M = 12,491 GWh
35% savings: 12,491 × 0.35 = 4,372 GWh
Homes equivalent: 4,372,000,000 ÷ 10,715 = 408,000 homes

Status: 🔶 DERIVED — Depends on 60M lights and 35% savings assumptions

When to Use This Skill

Use photometrics-financials for:

• •ROI calculations and proposal budgets
• •Value per light questions
• •Energy savings math
• •Market size and penetration data
• •Verifying any quantitative claim
• •Utility cost avoidance calculations
• •Pricing discussions

Use photometrics-ai (core skill) for:

• •How the product works
• •Technical architecture (TLL, optimization engine)
• •Competitive positioning
• •Go-to-market strategy
• •Use cases and applications