← Back to the plan

TRC Electrical Load Inventory, 1660 9th St, Santa Monica

Compiled 2026-09-06 from Greg's local files. Every row cites a source file. Rows with no nameplate source are marked ESTIMATE and carry a stated basis.

0. THE HEADLINE: measured 15-minute demand data already exists on this machine

The email dossier records that SCE owed TRC "13 months of 15-minute demand CSVs for both meters" and that none appear delivered. They were delivered, and they are on disk.

~/Library/CloudStorage/Dropbox-RecordingClub/Greg Spero/Greg's Dropbox/SCE/FULL DATA 2025-JAN2026/
  SCE_Usage_[account]_01-01-25_to_01-29-26.csv   (37,724 intervals)
  SCE_Usage_[meter 3]_01-01-25_to_01-29-26.csv   (37,724 intervals)

15-minute kWh "Delivered" per meter, 2025-01-01 through 2026-01-28. Demand kW = kWh x 4. Meter resolution is 0.4 kWh, so every reading quantizes to a multiple of 1.6 kW.

There are TWO service accounts at this address, not one

The March 2025 bill (TRC/Electric/ViewBill03262025.pdf, and the SCE folder bills) lists service accounts [account] and SCE meter 3, both at 1660 9TH ST, under customer account [account].

Source: SCE/ViewBill01052026.pdf lines for both accounts; CSV analysis above.

SCE meter 3 has never come close to 20 kW. The entire demand problem lives on [account] (the account the BRIEF says closed 1/29/26 and was replaced by [account]).

Measured monthly maximum 15-minute demand (kW)

Combined percentiles across all 37,724 intervals: p50 6.4 kW, p90 14.4 kW, p99 22.4 kW, p99.9 28.8 kW, max 36.8 kW. Overnight (3-5am) combined base load: median 3.2 kW, p90 6.2 kW.

The EV charger was the dominant driver on meter 754, and limiting it worked

Sustained runs at >=11 kW lasting >=1 hour, and intervals >=16 kW, on meter 754:

That step change lands exactly on the Nov/Dec 2025 EV-charger limiting recorded in context/email-dossier.md:16. Longest single sustained run: 660 minutes (11 hours). Greg measured Tesla charging at 11.3 kW (context/texts-ari-electrical.txt:161).

This contradicts the soft-starter theory on the physics

The billed quantity is a 15-minute average, not an instantaneous peak. The 2025-08-23 peak was 27.2 kW on meter 754 held across consecutive intervals from 16:45 to 18:45, and 2025-07-31 held 25.6-27.2 kW across 18:30 and 19:00. A compressor inrush lasting 0.5-3 seconds contributes at most 3/900ths of a 15-minute window. To move that average by 1 kW an inrush would have to last 5 minutes. The measured peaks are sustained running load, not startup transients. This corroborates Farzad (texts-ari-electrical.txt:772) and Victor (texts-ari-electrical.txt:976) against Ari's inrush theory. Confidence: high.

1. Load inventory

Phase note: Farzad, 2026-09-04, "sauna is the only equipment that uses 3 phase and there's already an existing 240V outlet. All the other work was done in 110V" (context/texts-taji-chats.txt:456). He then confirmed the AC units are 220V (texts-taji-chats.txt:458-460). Both statements can be true: the ACs are 240V single-phase (split-phase, two hot legs), which is what the one nameplate we have says.

HVAC, 7 rooftop package units

The building has 7 rooftop units total (TRC/AC/RFP/TRC_Live_Room_HVAC_RFP.md, section 2; TRC/AC/RFP/TRC AC Issue Document 3_18_26.pdf, "The building has 7 units total"). Only two have a documented model.

Corroboration on the count: Elite Heating's Apr 2025 estimate prices preventive maintenance on "the six HVAC rooftop systems" plus the one Lennox replacement = 7 (TRC/AC/Elite Heating and Air Conditioning - estimate_16393.pdf). That estimate also quotes a Lennox LRP13HPK60EP explicitly as "5 Ton, 208-230v 1ph 60 hz", further evidence the rooftop fleet is single-phase.

Also on record: the Live Room unit is starved on airflow, measured 1,528 CFM actual against 2,000 design (76%), total external static 1.44 in. w.c. against 0.8 rated (TRC/AC/Airflow Report - Recording Club.pdf, 04/16/2026, NCI cert 17-399-01). A unit fighting nearly double its rated static draws more power, not less. That is a live contributor to running demand.

HVAC subtotal, all 7 running: 24-33 kW ESTIMATE.

Non-HVAC loads

Panel infrastructure

• Main supply panel is in an electrical room inside the building, fed by an overhead service. Service amperage rating recorded as "200 / 400", the inspection report does not resolve which. Source: TRC/Inspection/Spero & Midler 1660 9th St. Santa Monica, CA 90404-Secured.pdf, "MAIN PANEL SPEC'S" and "PATH OF ELECTRICAL".
• Zinsco panels present. Inspection flagged them; insurance carrier waived replacement 2026-08-20. Expected 5-year electrical cost $10,000-$15,000+. Same report, ELECTRICAL section.
• AC units are on dedicated circuits wired directly to the roof panel (texts-ari-electrical.txt:145, :159).
• Farzad's crew ran a neutral wire back to the outdoor panel in Nov 2025 because outlet circuits lacked one (texts-ari-electrical.txt:514-517).

2. Scenarios

Assumptions common to both

1. Demand is a 15-minute average, so startup inrush is excluded from every figure below.
2. AC running kW uses total unit draw (compressor + condenser fan + indoor blower).
3. The Live Room unit runs above catalogue draw because it is fighting 1.44 in. w.c. against a rated 0.8. A 10% uplift is folded into its range.
4. Load allocation between the two meters is unknown. Both scenarios are stated facility-combined; the per-meter split is the single biggest open question, because the tariff threshold is per-meter.
5. The 7 rooftop units do not all run continuously. On a hot evening 5 of 7 running with 2 cycling gives a coincidence factor around 0.8.

Scenario A, worst-case coincident, hot day, evening session

Everything TRC can turn on at once, all seven ACs loaded, sauna hot, wall bright, car charging.

Sanity check against measurement: the highest 15-minute reading ever recorded across both meters is 36.8 kW (2025-08-23 17:45). That peak almost certainly did not include the sauna and the EV charger and a fully lit LED wall simultaneously, which is why the bottom-up worst case sits above it. Treat 45-55 kW as the true worst case and 36.8 kW as the demonstrated worst case. Confidence: moderate on the bottom-up figure, high on the 36.8 kW.

Scenario B, typical busy evening

Summer weekday, 6-9pm, a session in the Live Room and one other room, four ACs running, no sauna, no EV charging, LED wall on typical content.

Sanity check against measurement: combined 95th-percentile demand in summer evening hours is 22-24 kW (18:00-21:00, Jun-Sep). Winter evening 95th percentile is 16-18 kW. The bottom-up 27 kW sits just above the measured 95th percentile and below the measured max, which is the right place for it. Confidence: high.

What this means for the 20 kW ceiling

• The ceiling applies per meter, and only meter [account] / [account] has ever breached it. SCE meter 3 topped out at 14.4 kW over 13 months.
• On meter 754, the post-EV-limit record is: Nov 14.4, Dec 12.8, Jan 19.2 kW, winter only. There is no summer month on record with the EV charger limited. The data ends 2026-01-28. Ari's own texts then report 17.6 kW on 2026-02-28 and the dossier records a 20.8 kW spike around 2026-07-28.
• So the honest read is: the facility has not yet demonstrated a summer under 20 kW, and a PCOC declaration promising never to exceed it is not currently supported by evidence. The evidence does support that EV-charger control, not soft starters, is what moved the number. Confidence: high on the diagnosis, high on the gap.

3. What could NOT be determined, the on-site read list for Taji

Everything below is a physical nameplate read or a panel trace. None of it exists in any file on this machine.

A. Per rooftop unit (all 7), photograph the data plate on each

For each unit, in order, and tag each unit with the number staff already use ("units 4 & 7"):

1. Brand and full model number
2. Serial number
3. Nominal tons / BTU (often encoded in the model, e.g. the "060" in PKHP-M-0601BB = 5 ton)
4. Voltage and phase, the plate reads e.g. "208-230V / 1 Ph / 60 Hz". Confirm no unit is 3-phase. Victor believes 1 or 2 systems might be 3-phase (texts-ari-electrical.txt:980); Farzad says only the sauna is (texts-taji-chats.txt:456). One of them is wrong and the answer decides whether hard-start kits can be fitted at all.
5. RLA (rated load amps, compressor)
6. LRA (locked rotor amps, compressor), this is the inrush figure
7. MCA (minimum circuit ampacity) and MOCP / max fuse or breaker
8. FLA of the condenser fan motor and the indoor blower motor
9. Which breaker in which panel feeds it, and the breaker's amp rating
10. Which of the two SCE meters that panel is fed from , this is the highest-value single answer on the list

B. Sauna

• Heater brand, model, kW rating, voltage, phase
• Confirm it is genuinely 3-phase, or whether it is 240 V single-phase on a 3-wire feed that someone called "3 phase"
• Breaker amp rating and which panel
• Whether it has a timer or thermostat that can be scheduled off during peak hours

C. LED wall

• Number of power supplies (PSUs) behind the 49 cabinets, and the printed wattage on one
• The breaker(s) and circuit(s) feeding the wall, and their amp ratings
• Whether a brightness cap is set in the NovaStar VX600 (650 cd/m2 nominal; running it at 40% cuts wall draw roughly in half)

D. EV charger

• Brand, model, and the amperage the charger is currently limited to (this was changed in Nov/Dec 2025 and the new setting is not recorded anywhere)
• Breaker rating and which panel and which meter

E. Panels and meters

• How many meters and which panel each feeds. The bills show two service accounts. Taji should photograph both meter faces and record the service account number on each.
• Main service size: 200 A or 400 A? The inspection report writes "200 / 400" without resolving it.
• A photograph of every panel directory (the label card inside the door), so loads can be mapped to meters without guessing.
• Which panel the Zinsco breakers are in, and what they feed.

F. Remaining unknowns with no owner

• Water heater: gas or electric, and if electric its kW.
• Cold plunge chiller: brand, model, running amps.
• Kitchen: how many mini fridges, and where they are.
• Whether the Carrier 48ESNA4809030 (System #3) was actually replaced in 2025 or is still in service, the Reyes proposal exists, the acceptance signature line is blank, and separate invoices from Reyes, Pacific AirCo and Elite all sit in the same folder for what may be overlapping work.

4. Sources consulted

Searched and found nothing relevant: ~/Documents, ~/Desktop, ~/Projects, the Gregbrain Obsidian vault, ~/trc-inventory-db, ~/trc-manual, ~/trc-canonical.