Home Energy Storage Performance for Townhouses

Fourteen attached units, one pad-mounted transformer, eleven identical 100 A services. That row of townhouses is why I stopped reusing my detached-house sizing spreadsheet. The homeowner in unit 9 wanted whole-home backup for a basement sump pump, a furnace blower, and a home office. The battery was the simplest part of the job. The party wall, the 60-inch side passage, the shared transformer secondary, and an HOA architectural committee were the real constraints. home energy storage performance for townhouses is governed less by the pack datasheet than by the geometry and electrical topology of attached housing — and that is what this guide covers.

Townhouses sit in an awkward middle: legally single-family dwellings with individual meters, but physically attached and often sharing a utility transformer with five to eight neighbors. Once you accept that, the design work becomes measurable instead of aspirational.

Why the Detached-House Playbook Fails on Attached Units

On a detached house I have four sides of envelope, a garage wall, an unconditioned basement, and typically 200 A of service. A townhouse gives me two usable elevations, a passage sometimes only 36 to 60 inches wide, a rated separation on both sides, and a service that is often 100 A or 125 A because the building predates electrification of everything.

• Working clearance wins over preference. NEC 110.26 requires 36 inches of depth in front of equipment likely to be examined while energized, with a 30-inch minimum width and 6.5 ft of headroom. In a 48-inch passage, that math decides the layout before aesthetics or HOA preference gets a vote.
• The service, not the battery, sets the ceiling. A 100 A busbar rarely accepts more than a 20 A backfeed without changing the interconnection method. That is 4.8 kW continuous — enough for a townhouse base load, not for a heat pump plus range plus EV.
• Shared secondaries amplify everything. When eight units hang off one 50 to 75 kVA pad-mounted transformer, one neighbor’s array plus another’s charging can push service voltage toward the ANSI C84.1 Range A ceiling of 126 V (252 V line-to-line) on a summer afternoon.

I measure before I model. A two-week interval logger on the main costs almost nothing and replaces a great deal of guessing. On that row, base load ran 0.38 kW at 3 a.m. and peaked at 7.1 kW at 6:40 p.m., with each unit averaging 38 to 44 kWh/day. Those two numbers define the useful envelope of any home energy storage system I specify for attached housing.

Sizing a Battery Against a Real Townhouse Load Profile

The most common townhouse quoting error is sizing to annual consumption instead of the load curve. A three-bedroom unit using 1,240 kWh per month draws roughly 41 kWh/day, but between 11 p.m. and 6 a.m. it drew only 17 to 20 kWh. That is the number a backup or time-of-use battery has to carry.

• Backup sizing: multiply overnight base load by the autonomy you actually promise. A 13.5 kWh LFP pack with 90 percent usable energy delivers about 12.1 kWh, covering 12 to 16 hours of that overnight load with margin. Promising 24 hours of whole-home autonomy on a 100 A service is a sales claim, not an engineering one.
• Peak shaving sizing: take the four highest 15-minute intervals of the billing month, subtract the target demand, and multiply by duration. Shaving 3 kW for two hours needs 6 kWh of usable throughput plus conversion losses.
• Throughput check: divide annual shifted kWh by 0.88 round-trip efficiency, then compare against the warranty throughput cap. A pack under a 10-year/70-percent capacity warranty with a 40 MWh throughput limit is often throughput-limited before it is cycle-limited in a two-cycle-per-day duty.

I also add a thermal allowance. Shaded wall cabinets in my region ran 5 to 9 °C above ambient in summer, and an exterior wall in January dropped the pack to 2 to 6 °C. LFP cells should not accept charge below roughly 0 °C, so cold-weather autonomy claims need a 60 to 110 W heater load added to the night budget.

Service Size, Busbar Math, and the 120 Percent Rule

Interconnection method is where townhouse projects get expensive. NEC 705.12 permits supply-side or load-side connection, and the load-side rule most installers reach for is the 120 percent busbar allowance:

• Maximum backfeed = (1.2 × busbar rating) − main breaker rating.
• 100 A busbar, 100 A main: 120 − 100 = 20 A, or 4.8 kW at 240 V.
• 200 A busbar, 200 A main: 240 − 200 = 40 A, or 9.6 kW.
• 225 A busbar, 200 A main: 270 − 200 = 70 A — the classic reason to order the larger enclosure even on a 200 A service.

Older rowhouse stock rarely has that headroom, so I use one of two alternatives:

• Supply-side tap, NEC 705.11. Land the output between the utility service point and the main disconnect with a properly rated, labeled disconnect. On the fourteen-unit row this avoided a service change entirely: roughly $800 to $1,500 for the enclosure and disconnect against $4,000 or more for a 100 to 200 A upgrade with utility coordination.
• Power control system, NEC 705.13. A listed PCS with UL 1741 SB certification and current transformers limiting export in real time removes the 120 percent arithmetic from the permit review. I default to this when the service is undersized but the owner still wants a 9.6 kW-class inverter. Expect a $300 to $600 adder for the gateway and CTs.

Siting: Party Walls, Side Yards, and Fire Separation

Townhouses are separated by rated assemblies — in the row I surveyed, two-hour walls tested to the ASTM E119 fire-resistance methodology. That rating is what the building official protects, so my siting rules for attached housing start there.

• Do not mount on a shared wall. Keep the rated assembly intact. Mount on the unit’s own exterior elevation, on a pad with a listed enclosure, or on an interior wall that is not part of the separation assembly.
• Treat indoor capacity limits seriously. Model codes cap unrated indoor residential installations on the order of 20 kWh per dwelling unit, with separation, sprinkler, and aggregate allowances that vary by edition. A 13.5 kWh cabinet fits; a 30 kWh stacked pair may not.
• Match the enclosure to exposure. NEMA 3R is the floor for weather; coastal and salt-fog sites get NEMA 4X or IP66 stainless. Every penetration is a future leak, so I specify bottom-entry glands with drip loops.
• Respect setbacks and passage width. Keep clear of the property line, the adjacent unit’s egress path, and the snow-shedding zone below a roof plane. Passages under 48 inches push the design to a rear pad.
• Expect HOA architectural review. On that row the committee approved the cabinet only after I supplied a dimensioned elevation, a color match to the trim, and a one-page note on setback and noise. Submitting it up front saved three weeks of delay.

Shared Transformer Secondaries and Voltage Rise

This is the townhouse failure mode most often misdiagnosed as an inverter fault. Eight attached units behind a 75 kVA pad-mounted transformer, four with 6 to 8 kW arrays, measured 249 to 253 V at midday. Under IEEE 1547 ride-through settings one inverter tripped, and the homeowner blamed the battery.

PV export raises voltage at the point of common coupling; battery discharge does nothing to help, while battery charging is the only load that pulls it back down. ANSI C84.1 Range A allows 114 to 126 V at the 120 V service point — generous until a shared secondary concentrates generation into one afternoon window. My mitigation sequence:

• Move charging into the low-load window between 10 p.m. and 6 a.m., which also matches most time-of-use rate structures.
• Enable volt-var and voltage-watt functions so the inverter stops exporting before the upper limit is reached.
• Ask the utility to verify transformer loading and secondary conductor sizing; the secondary is sometimes the constraint, not the inverter.
• Where adjacent units both add storage, coordinate commissioning so both systems are not exporting at full power simultaneously.

Batteries genuinely help here, but only when configured to charge during the solar peak. One townhouse pair I commissioned held both services under 250 V all summer with coordinated morning charging.

Measured Performance: Efficiency, Derating, and Noise

Datasheet performance and townhouse reality differ in four measurable places. Over twelve months on that row I recorded:

• Round-trip efficiency: 87 to 92 percent measured DC-to-AC across a full cycle, including both conversion stages and standby. The headline 95 percent figure is a DC-side cell number, not a system number.
• Standby drain: 15 to 40 W continuous housekeeping load, or 130 to 350 kWh per year — roughly $20 to $55 annually before any shifting benefit.
• Thermal derating: a west-elevation cabinet in direct sun reached 48 °C internally and the BMS reduced charge current; shaded or north-side siting restored full rate. Winter capacity at 5 °C was 88 to 92 percent of rated before heater draw.
• Noise: 45 to 55 dBA at one meter under load. That matters in a three-foot passage beside a neighbor’s bedroom window when the local night limit is 45 to 55 dBA at the property line. A rear pad or a quieter fan curve solves it.

Cycle-life claims also need translation. A 6,000-cycle LFP rating at 25 °C and 80 percent depth of discharge is credible, but cycles are consumed by throughput: a townhouse running two cycles daily for ten years accumulates roughly 7,300 cycles. That is why I compare warranty throughput caps and temperature history rather than the marketing cycle count.

The Specification and Commissioning Checklist I Use for Townhouses

• Two weeks of 15-minute interval data from the unit’s own meter, plus a two-day logger on the shared secondary for voltage.
• Busbar and main breaker ratings photographed with the cover off — nameplate assumptions are wrong on roughly one in three older rowhouses.
• Interconnection method confirmed in writing, and siting documented with side-yard dimensions, working clearance, party wall location, and HOA approval attached.
• Enclosure rating matched to exposure, with bottom-entry glands and drip loops.
• Certifications verified for the shipped configuration: UN38.3 transport testing, IEC 62133 cell and pack safety, UL 9540 for the energy storage system, and UL 9540A data where the code official requests thermal runaway propagation evidence.
• Commissioning record covering measured round-trip efficiency over a full cycle, standby load, peak internal temperature at full power, and property-line sound level.
• Handover sheet with cold-weather behavior, charge window settings, and a realistic autonomy number for each backed-up circuit.

None of this is exotic. The difference between a decade of quiet performance in a 100 A townhouse and a warranty call every winter is rarely the pack.

Frequently Asked Questions

Can a townhouse with a 100 amp service support home energy storage?

Yes, within limits. A 100 A busbar with a 100 A main yields a 20 A load-side backfeed allowance under the 120 percent rule, about 4.8 kW continuous. That covers base loads, a furnace blower, a sump pump, and a home office, but not a heat pump, range, and EV charger at the same time. Many of my townhouse projects go further with a supply-side tap per NEC 705.11 or a listed power control system with UL 1741 SB certification, neither of which required a service upgrade.

How much battery capacity does a townhouse actually need for backup?

Size to the overnight base load, not annual consumption. In the attached units I metered, overnight base load ran 17 to 20 kWh between 11 p.m. and 6 a.m., against 38 to 44 kWh of daily consumption. A 13.5 kWh LFP pack with about 12.1 kWh usable covers 12 to 16 hours with margin. If you want 24 hours of whole-home autonomy you are specifying 30 to 40 kWh plus a service review, which most 100 A townhouses cannot supply continuously anyway.

Where can I install a home energy storage system on an attached house with no garage?

Three locations, in the order I try them: an exterior wall on the unit’s own elevation, a pad at the rear or in a side passage with the required working clearance, or an interior wall that is not part of the separation assembly. Avoid the rated party wall. Passages narrower than 48 inches usually fail NEC 110.26 clearance once cabinet depth is added, which pushes the design to a rear pad. Coastal sites need NEMA 4X or IP66 rather than the minimum NEMA 3R.

Does a townhouse need a permit or HOA approval for residential battery storage?

Both, in most jurisdictions. The permit package typically needs the interconnection method, busbar and main breaker ratings, an equipment listing, and sometimes thermal runaway propagation documentation such as UL 9540A data. Separately, HOA architectural review controls exterior appearance, placement, and color. On the row I worked, a dimensioned elevation, a trim color match, and a one-page setback and noise note cut about three weeks of delay versus a verbal request.

What fire separation is required between a battery and a shared party wall?

Townhouses are separated by rated assemblies, commonly two-hour walls tested to the ASTM E119 methodology, and that rating must remain intact. I do not mount storage equipment on the separation assembly, and I maintain listed clearances to it. Unrated indoor installations are also capped by model codes on the order of 20 kWh per dwelling unit, so larger banks often go outside in a listed enclosure.

Will a home battery backup help with voltage rise on a shared transformer?

It can, if charging coincides with the solar peak. I measured 249 to 253 V at midday on one eight-unit secondary with four arrays, at the edge of ANSI C84.1 Range A. Discharging does not lower voltage; charging does. Setting the battery to absorb surplus between 10 a.m. and 2 p.m., with volt-var and voltage-watt enabled on the inverter, kept both services under 250 V through the summer without curtailing the arrays.

How much capacity does a home energy storage system lose in winter on an exterior wall?

Expect usable capacity near 88 to 92 percent of rated at 5 °C, and less below freezing because LFP cells should not accept charge under roughly 0 °C. Heated enclosures preserve charge acceptance but add 60 to 110 W of continuous load, which belongs in the night autonomy budget. My winter measurements on a shaded exterior wall showed pack temperatures of 2 to 6 °C, so I either specify a heated enclosure or move the unit into conditioned space where the code permits it.

What is the realistic round-trip efficiency of a home energy storage battery in a townhouse installation?

I measure 87 to 92 percent DC-to-AC across a full cycle, including both conversion stages and parasitic housekeeping load. Datasheet figures near 95 percent usually describe the DC-side cell and exclude the inverter, gateway, and standby consumption. Standby alone runs 15 to 40 W, or 130 to 350 kWh per year. When I model time-of-use savings I use 88 percent and a documented standby draw rather than the marketing number.