Scope 2 Utility Data Pipeline Starter Kit

Build a defensible kWh dataset (with billing dates + evidence) that doesn't collapse under scrutiny

Carbon BasicsCoreStarter Kit1 hour

What you'll accomplish

  • Build a portfolio utility dataset with consistent columns and evidence links
  • Stop the "we can't get the data" cycle with a clear collection workflow
  • Handle mixed reality: owner-paid, tenant-paid, master meters, submeters
  • Produce Scope 2-ready inputs (kWh + billing periods) with a simple QA process
  • Create an operating cadence: monthly closeout + exception handling

Who this is for

Sustainability teams

Responsible for carbon reporting

Finance/Ops teams

Who want clean, auditable utility spend and usage

Property management & leasing teams

Who touch tenant utility access

Anyone

Building carbon fundamentals for real estate

When to use this

Use this when:

  • •You're starting Scope 2 for the first time
  • •You have incomplete or inconsistent utility bills/portal exports
  • •You need evidence links for every data point
  • •Your portfolio has tenant-paid utilities and you don't have a system

Prerequisites

Start Here + Glossary (kWh vs kW, billing periods)Utility Responsibility in Real Estate (recommended)Contract Repository Setup (optional, to store evidence consistently)

Quick start (60 minutes)

Pilot with one building and one utility type (electricity):

  • Create the Utility Roster (Template 1) for that building
  • Create the Intake Table (Template 2)
  • Collect 2–3 utility bills or a portal export
  • Populate: billing start/end dates + kWh + account/meter ID
  • Add an evidence link for each record (PDF/export)
  • Run QA checklist (Template 4)
  • Schedule monthly closeout (Template 5)

The minimum viable Scope 2 dataset (what "done" looks like)

For each site/account/billing period you need:

Site / building identifier
Utility type (electricity, gas, etc.)
Meter/account ID
Billing period start date
Billing period end date
Usage (kWh for electricity)
Unit (kWh)
Data source (bill PDF, portal export)
Evidence link (where the proof lives)
Data quality label (Measured / Allocated / Proxy)

Beginner rule: If you have kWh + dates + evidence, you can build from there.

Step-by-step pipeline

1

Build the Utility Roster (the map)

Your roster answers: Which sites exist? Who pays (owner vs tenant)? Where does data live (portal, tenant bills, PM records)? What's the collection pathway?

Done looks like: Every site has an assigned pathway and owner.

2

Choose your evidence storage pattern

Pick one canonical place for evidence: One folder per site (recommended), within site: utility type folders, within utility type: YYYY/MM statements.

Done looks like: You can click from dataset row → evidence in one step.

3

Collect data using the lowest-friction method first

Preferred order: 1) Utility portal export (best for consistency), 2) Bill PDFs (good, requires manual extraction), 3) Tenant-provided bills (common in NNN), 4) Metering platform exports (if installed), 5) Proxy (temporary only).

Done looks like: You can consistently collect at least one method per site.

4

Normalize the dataset (make columns consistent)

Standardize: Units (kWh), Dates (YYYY-MM-DD), Site naming (single canonical ID), Billing period definitions (do not assume calendar months).

Beginner trap: Billing periods rarely match calendar months.

5

Quality assurance (QA)

QA checks: billing periods don't overlap, no missing dates, kWh is not confused with kW demand, totals are reasonable vs prior periods, evidence links exist and open.

Done looks like: Dataset is defensible, even if incomplete.

6

Closeout cadence (monthly)

At month-end: ingest new bills/exports, update dataset, run QA, log exceptions (missing data, tenant nonresponse), escalate systematically.

Templates included

Template 1 — Utility Roster (copy/paste table)

| Site ID | Site name | Address | Utility type | Paid by (Owner/Tenant/Mixed) | Meter/account ID known? | Data source (portal/bills/tenant/metering) | Data owner | Collection pathway | Notes |
|---|---|---|---|---|---|---|---|---|---|

Template 2 — Utility Intake Table (Scope 2 minimum viable)

| Site ID | Utility type | Account/Meter ID | Billing start | Billing end | Usage | Unit | Source type | Evidence link | Data quality (Measured/Allocated/Proxy) | Notes |
|---|---|---|---|---|---:|---|---|---|---|---|

Template 3 — Data request email (tenant or landlord)

Subject: Utility usage data request (kWh) for [Building] — [Period]

Hi [Name],
We're building a verified utility usage dataset for [building/portfolio]. Can you provide electricity usage for:

Building / Suite:
Billing period(s):
Preferred: utility portal export or bill PDFs showing kWh and billing dates.

If you can, please include:
- Billing period start/end dates
- Total kWh
- Account/meter identifier
- PDF or export file (evidence)

Thanks — once we standardize the format, future requests will be quick.
[Name]

Template 4 — QA Checklist (copy/paste)

Utility Data QA Checklist

Required fields
- Site ID present
- Billing start and end dates present
- Usage present (kWh for electricity)
- Unit correct (kWh)
- Evidence link present and opens

Consistency checks
- No overlapping billing periods for same site/account
- Billing periods are continuous or gaps are logged as exceptions
- Usage is within a reasonable range vs prior periods (flag spikes)

Common errors
- kW (demand) entered instead of kWh
- Calendar month assumed instead of billing period
- Multiple meters combined without noting it

Template 5 — Monthly Closeout Ritual (30–60 minutes)

Monthly Utility Closeout

1) Collect new bills/exports (20 min)
- Owner-paid accounts
- Tenant-paid accounts (requests + reminders)

2) Update intake table (20 min)
- Add rows for new billing periods
- Add evidence links

3) QA (10 min)
- Run checklist
- Flag exceptions

4) Exceptions + escalation (10 min)
- Missing tenants
- Missing portals
- Data gaps
Assign owners + dates

Template 6 — Exceptions Log (copy/paste)

| Site ID | Utility type | Period missing | Reason | Owner | Next action | Due date | Status |
|---|---|---|---|---|---|---|---|

Common pitfalls

  • ✗Treating "kWh" as optional (it's the foundation)
  • ✗No evidence links (kills credibility)
  • ✗Mixing site naming conventions (breaks analysis)
  • ✗Not tracking billing periods (creates overlaps/double counting)
  • ✗Leaving tenant data to "PM follow-ups" with no escalation ladder

How to prove impact

% of sites with kWh + dates + evidence links

Track coverage

Data latency (days from billing end to dataset updated)

Should decrease

Reduction in missing periods over time

Gaps should shrink

% measured vs allocated vs proxy

Proxy should shrink

Evidence and Confidence

Confidence:High(data pipeline principles are stable)

Assumptions: You can access some utility data and enforce evidence storage.

Where this can fail: If tenant-paid utilities are unmanaged and escalation is missing.

Change log

v1.0 (2026-01): Latest release