This group brings together the best thinkers on energy and climate. Join us for smart, insightful posts and conversations about where the energy industry is and where it is going.


Comparison of Grid-Connected and Off-the-Grid Houses

Willem Post's picture
President Willem Post Energy Consuling

Willem Post, BSME'63 New Jersey Institute of Technology, MSME'66 Rensselaer Polytechnic Institute, MBA'75, University of Connecticut. P.E. Connecticut. Consulting Engineer and Project Manager....

  • Member since 2018
  • 613 items added with 219,559 views
  • Nov 26, 2014

grids and houses

National Energy Systems are Wasteful: About 75% of the fossil energy taken out of the ground to generate electricity never reaches the end user as electricity due to various losses from mine or well to user’s meter, and due to changes in embedded energy due to repairs, replacements, enhancements, expansion, etc., of the various systems, from mine or well to meter, plus, for a more inclusive approach, the energy required for the various other activities of the power industry-government complex. 

Then the user feeds the energy into devices with efficiencies as little as 5%, such as incandescent light bulbs. Such wasteful, national energy systems developed over the decades, because the fossil energy was, and still is, low-cost and abundant.

Even district CHP plants with thermal and electrical distribution systems, popular in Denmark, etc., have low efficiencies, if all energy losses and costs are taken into account; the CHP plants may be efficient, but the distribution systems and the mostly energy-hog buildings usually are not.

If the buildings had been designed to the Passivhaus standard, their energy consumption for heating, cooling and electricity would have been about 4 times less, and the capacity of the CHP plant and distribution systems, and of the energy from mine or well, and CO2 emissions would have been about 4 times less!

Building Energy Efficiency 

As all technologies are fully developed and proven, more energy could be locally generated and locally consumed in energy-efficient buildings, all “under one roof”, as shown by the alternatives in the article. There would be massive resistance from special interests to go into that direction, as they have grown big by exploiting the fossil fuel-addicted society for at least the past 100 years.

The energy efficiency of buildings did not become an issue until after the 4-fold increase of crude oil prices in 1973. The owners of mostly energy-hog buildings, seeing major increases in their heating and cooling costs, consulted with engineers to make energy surveys of buildings, which, after implementation of the recommendations, usually resulted in at least 50% decreases of energy consumption.

Such efficiency improvements regarding houses did not take place until much later, and then only on a case by case basis, because politicians were, and still are, very slow to upgrade building codes. For them, it is so much easier to cater to special interests, to be for heavily subsidized, highly visible, renewable energy, than for lightly subsidized, invisible, energy efficiency.

Because CO2 emissions are one of the factors affecting global warming and climate change, it would be desirable to have new buildings be near the goal of “net-zero-energy and near-zero CO2 emissions”.

Typical Air Leakage of a Free-Standing House

Typical leakage rates of 2,000 ft2 houses (16,000 ft3 volume) , as determined by blowerdoor tests, are:

– Older house…………..2133 ft3/m……………….8.0 ACH @ 50 Pa, or greater

– 20009 IECC…………………………………………..<7 ACH @ 50 Pa for Climate Zones 1 and 2

………………………………………………………………<5 ACH @ 50 Pa for Climate Zones 3 – 8

– New house……………1040 ft3/m……………….3.9 ACH @ 50 Pa, or greater*

– 2012 IECC…………………………………………….<5 ACH @ 50 Pa for Climate Zones 1 and 2*

………………………………………………………………<3 ACH @ 50 Pa for Climate Zones 3 – 8*

– Tight house…………….400 ft3/m……………….1.5 ACH @ 50 Pa, or less*

– Passivhaus……………..160 ft3/m……………….0.6 ACH @ 50 Pa, or less*

* These leakages would be significantly less at lesser pressures. A ventilation system with heat-recovery ventilator, HRV, would be required.

Space Heating Demand

Typical space heating demands of 2000 ft2, freestanding houses are:

House type…………………………………… Heating Demand…………Heating System Capacity


 – Older house, in New England……………….45 – 55…………………….up to 125,000  

– Newer house in New England………………..20 – 25 …………………….up to  60,000  

– Tight house in New England…………………….8.5…………………………………17,000; see NOTE.

– Passivhaus…………………………………………… 3.2………………………………….6,348; see NOTE.  

NOTE: Here is the URL of a 1,232 ft2 tight house with a PEAK space heating demand of 10,500 Btu/hr, or 8.5 Btu/ft2/h, an equivalent 2000 ft2 house would have a PEAK space heating demand of about 2000/1232 x 10500 = 17,045 Btu/hr. It uses for:

– Heating: two Mitsubishi, Mr. Slim, ductless, minisplit, heat pumps (one downstairs @ 12,000 Btu/hr, and one upstairs @ 9,000 Btu/hr), installed cost about $5,250.

– Ventilation; a Lifebreath 155 ECM energy-recovery ventilator.

– Electricity: a grid-connected, PV system, 5.7 kW, roof-mounted with Fronius IG 5100 inverter, installed cost about $22,000 less subsidies.

NOTE: Passivhaus is about 10 W/m2 x 186 m2 = 1.86 kW, or 6,348 Btu/hr, or 3.2 Btu/ft2/h, i.e., a 2 kW, thermostat-controlled, electric heater in the air supply duct COULD be the heating system!!

Reducing Heat Loss of a Free-Standing House

The heat loss is given by Q = UA x (Tref – Toutside), where Q is in Btu/h, U is in Btu/h/ft2/F, A is in ft2, T is in degree F. The UA value, in Btu/h/F, a property of the building envelope, can be determined by summing the UA values for each wall of the basement, house and roof.  A straight-line graph connecting the plotted values of recorded heating system inputs to the building, Btu/h, over several years, versus outdoor temperature, F, would have UA as the slope of the line. 

A large, older, poorly insulated house (high U-values and large areas), heated the traditional way, may have a 4 to 5 times steeper slope than a small, energy-efficient house heated with heat pumps.

NOTE: A mini-split, cold-climate heat pump may have a rated COP of 3.0 at 47F exterior temperature; heat pumps are rated at 47F, per industry agreement. The COP, decreasing with exterior temperatures, may be as follows: 1.5 at -15F (at maximum compressor speed/high fan speed, i.e., noisy); 2.0 at 0F; 2.5 at 32F; and 3.0 at 47F. See page 10, figure 5 of this URL. Operating a heat pump at a COP of 2.0 or less is likely more costly than a wood-fired stove, or a 95%-efficient propane-fired, condensing furnace, or a thermostat-operated, high-efficiency gas/propane stove (makes no noise). As a result, about 70% of space heating typically is provided with heat pumps, and the rest with conventional heating units during colder days.

The house thermostat may be set at 68F, but the heating system typically does not start operating, because of passive solar heat gains and internal heat sources (computer, lighting, cooking, people, etc.), which keep the house indoor temperature above 68F, even though the outdoor temperature is lower.

Due to those heat gains, the heating systems of energy-efficient houses (tight and Passivhaus) would not need to start operating, except on colder days, and supply only small quantities of heat. Invisible energy efficiency at work!!

The below descriptions of energy-efficient construction can be performed by existing skilled building trade personnel and by capable do-it-yourselfers without specialized tools or training. Blown-in cellulose and sprayed foam by special contractors are avoided.

R40 Wall: An R40 wall (2 x 4; cavity section) can be achieved as follows: 1 (surface film) + 0.45 (½” plasterboard, NO PVC vapor barrier; see NOTE) + 12.25 (3.5” fiberglass @ 3.5/in) + 0.63 (½” plywood sheathing) + 0 (6 mil PVC vapor barrier) + 20 (4” blueboard @ 5/in on OUTSIDE of sheathing) + 0.81 (1/2” siding) + 1 (surface film) = 36.14; this value increases to about 1.1 x 36.14 = 39.75 at -10F. The blueboard is affixed to the house framing with 1 x 3 strapping and long screws. The siding is affixed to the strapping.

R48 Wall: For a 2 x 6 wall, with 5.5 inch of fiberglass, the R-values would be 43.14 and 47.45, respectively.

NOTE:  If 3” of blueboard were snugly fit into the cavities of a 2 x 6 wall, instead of fiberglass, the wall R-value would be 38.89 (42.78 at -10F), but there would be a 2.5” space for wiring and piping. 

NOTE: The R-value of 3.5” fiberglass increases from 12.6 at 75F to 14.5 at -10F, and of 4” blueboard, 44 months old, increases from 20 at 75F to 22 at -10F.

NOTE: The vapor barrier is on the outside of the sheathing, because its temperature would be almost always above the dew point. There must be NO vapor barrier between the sheetrock and the stud wall so the wall can breathe to the interior.

NOTE: In the above R40 and R48 walls, fiberglass insulation with air passing through it would act as a filter, instead of an insulator, if it were not for the 4″ of blueboard on the exterior of the sheathing.

R20 Basement: An R20 basement can be achieved as follows: two layers of 2″ thick x 2′ x 8′, 100 psi, blueboard (special order at Home Depot) under the 18″ wide concrete footing + 4″ of standard, 25 psi, blueboard under the basement slab, on the outside of the footing, outside of basement wall, outside of house sheathing, up to the roof overhang. With seams staggered and taped, such a foam enclosure will provide at least R20 everywhere, with no thermal bridges, and with near-zero air infiltration.

NOTE: The exposed blueboard must be covered with stucco or 1/2″ pressure-treated plywood that must to be stained at least every 5 years.

NOTE: The weight of a 2-story house, including concrete basement wall and footing, results in about 10 psi pressure on the 100 psi-rated blueboard under the footing.

NOTE: Multiply the U-value in standard international, SI, units, W/m2/C, times 0.1761 to obtain U-value in English units, Btu/h/ft2/F.

(3.4129 Btu/h)/10.76 ft2/1.8F = 0.1761

R65 and R96.5 Ceilings: An R65 ceiling can be achieved as follows: 1 (surface film) + 31.5 (9“ fiberglass between the 2 x 10 joists) + 31.5 (9” fiberglass across the joists) + 1 (surface film) = 65. The attic should have a 2-ft knee-wall to facilitate installing the insulation. On the OUTSIDE of the sheathing of that knee-wall are the vapor barrier and the 4” of blueboard. An extra foot of knee-wall would serve for the future addition of 9” fiberglass for an R96.5 ceiling!!

IECC Leakage Standards for Windows and Doors: IECC standards* for maximum fenestration air leakage are:

– Operable windows, skylights, and sliding glass doors: 0.34 ft3/m per lineal foot of operable sash crack, or 0.30 ft3/m per square foot of window area.

– Residential doors, swinging:…………………………………..0.50 ft3/m per square foot of door area.

– Residential doors, sliding:………………………………………0.37 ft3/m per square foot of door area.

* Standards are not requirements, i.e., double-hung windows may have leakage rates much greater than the IECC standard.

If a 2,000 house has 300 ft2 of windows, then the leakage rate could be 300 x 0.30 = 90 ft3/m on cold, windy days, a significant percentage of the leakage rates of tight and Passivhaus houses. Additional leakage occurs around the window, if the nailing flange of the window is not sealed with tape and the space between the rough opening and window is not properly sealed. If not taped and sealed, that leakage may be greater than the window leakage!

Windows: A vinyl-clad, wood-frame, double-glazed, argon-filled, low E window, without grilles, such as from Anderson, Pella, and Jen-Weld, may have a label stating a U-factor of 0.29, SHGC of 0.32, VT of 0.54, and air leakage* of 0.3. South-facing windows of passive-solar houses should have high SHGCs. 

* Air leakage is an optional rating, and manufacturers can choose not to include it on their labels.

High R-value windows, i.e., R5 to R7, usually have low leakage rates, but they are much more expensive than standard R3 (U-factor 0.33) windows, which are mass-produced. It is more cost-effective to spend additional effort to properly install standard windows.

Doors: The higher the R-Value of the energy-efficient entry door, the lower the heat loss. For example:

– Six-panel wood door, no window, 1.5 inch………………………………1.5

– Solid wood door, no window, 1.5 inch…………………………………….2.0

– Typical fiberglass door, no window, 1.5 inch…………………………..5 – 6

– Custom fiberglass door system, no window, 2 inch……………….10 – 12

Grid-Connected and Off-the-Grid Alternatives for Houses

Below are three energy, CO2 emission, and cost reduction alternatives for houses; the first two ones go only part way towards the goal, the last one goes much further.

Alternative No. 1 is having a standard, code-designed house, with a grid-connected PV solar system of sufficient capacity to charge a plug-in vehicle. This alternative would achieve CO2 emission reductions, but would be a long way off from “net-zero-energy and near-zero CO2 emissions”.

Alternative No. 1A is having a very energy-efficient, Passivhaus level house, with a grid-connected PV solar system of sufficient capacity to charge a plug-in vehicle. This alternative would come much closer to “net-zero-energy and near-zero CO2 emissions”.

Alternative No. 2 is the same as 1A, except it has suitable systems to be “off-the-grid”. This alternative would come much closer to “net-zero-energy and near-zero-CO2 emissions”, and, as a side benefit, would reduce the power industry-government complex, if widely adopted.

NOTE: There are 11 notes at the end of the article.

Alt. No. 1: Standard House, With Grid-Connected PV Solar System and Plug-in EV

This alternative is used worldwide, especially in Germany. Its main attraction is using the generators on the grid to supply steady, 24/7/365 energy when PV solar energy is insufficient or absent, at least 80% of the hours of the year.

In effect, the grid connection is a valuable, free (to the homeowner) energy service mostly paid for by the other ratepayers. To add to that free service, politicians often bestow high feed-in rates for any excess PV solar energy that cannot be used by the household. The house:

– Space heating and DHW consumption would be about 600 gal fuel oil x 138,500 Btu/gal = 83,100,000 Btu/yr.

– Electrical consumption would be about 6,000 kWh/yr, or 20,478,000 Btu/yr, equivalent to a 6,000 kWh/1,226.4 kWh/yr/kW = 4.9 kW PV solar system in New England.

– Plug-in EV consumption would be about 12,000 mi/yr x 0.30 kWh/mi = 3,600 AC kWh/yr, as measured to the charger, equivalent to a 2.9 kW PV solar system in New England

If the house is equipped with a 4.9 + 2.9 = 7.8 kW PV solar system, the house and plug-in electricity are offset; the plug-in is assumed to operate only on electricity.

Total Site Energy: Site energy, including the plug-in, would be 83,100,000 Btu/yr. Without PV solar systems it would have been 115,864,800 Btu/yr.

Investment and Energy Cost Savings: The cost of the PV system would be about 7.8 x $4,000/kW = $30,400, less subsidies. Without PV solar system, annual bills would be for electricity $0.18/kWh x 6,000 kWh = $1,080, and for plug-in 0.18 x 3,600 = $648. With PV solar system, they would be minimal, but bills for space heating and domestic hot water, DHW, about 600 x $3.50/gal = $2,100/yr. (about $2,800 before tax), would remain.

Alt. No. 1A: Passivhaus, With Grid-Connected PV Solar System and Plug-in EV

In New England, Note 8 implies the following minimum values: Basement R-20, Walls R-40, Ceiling (or roof) R-60. In colder climates, such as Canada, these values would be greater. The house:

– Space heating consumption would be about 15 kWh/m2/yr x 186 m2 = 2,790 kWh/yr, or 23.3 kWh/day, if averaged over 4 months. The house would consume for space heating = 2,790 kWh/yr x 3,413 Btu/kWh x 1 gal/138,500 Btu x 1/0.75 eff = 91.6 gallon of fuel oil, or 12,687,732 Btu/yr, if all the heating demand were met only with fuel oil. See Note 9.

– PEAK space heating demand would be about 10 W/m2 x 186 m2 = 1.86 kW, or 6,348 Btu/hr, or 3.2 Btu/ft2/h, i.e., a 2 kW electric heater in the air supply duct COULD be the heating system!!

– DHW consumption, 2 occupants, would be about 20 kWh/m2/yr x 186 m2 = 3,720 x 3,413 x 1/138,500 x 1/0.75 = 122.2 gallon of fuel oil, or 16,928,480 Btu/yr. Because of the energy efficiency of the house, the DHW energy became greater than the space heating energy, whereas in a standard house, it is about 20% of the space heating energy.

– Electrical consumption would be about 2,855 kWh/yr, or 9,742,504 Btu/yr, equivalent to a 2.3 kW PV solar system in New England. 

– Plug-in EV consumption would be about 12,000 mi/yr x 0.30 kWh/mi = 3,600 kWh/yr, as measured to the charger, equivalent to a 2.9 kW PV solar system in New England

If the house is equipped with a 2.3 + 2.9 = 5.2 kW PV solar system, the house and plug-in electricity are offset; the plug-in is assumed to operate only on electricity.

Total Site Energy: Site energy, including the plug-in, would be 29,616,212 Btu/yr. Without PV solar systems it would have been 51,645,516 Btu/yr.

Investment and Energy Cost Savings: The cost of the PV system would be about 5.2 x $4,000/kW = $20,000 less subsidies. Without PV solar system, annual bills would be for electricity $0.18/kWh x 2,855 kWh = $514, and for plug-in 0.18 x 3,600 = $648. With PV solar system, they would be minimal, but bills for space heating and DHW, about 213.8 gal x $3.50/gal = $748/yr. (about $1,000 before tax), would remain.

Alt. No. 2: Energy Efficient House, Off the Grid, With PV Solar System and Plug-in EV

This alternative is becoming increasingly attractive, as the prices of PV solar systems have decreased and subsidies are generous. As battery systems become more widely used for electrical energy storage, their prices will decrease as well. Homeowners should receive the same 30% subsidy for the battery systems, as now applies to PV solar systems. The off-the-grid mode is distributed mode energy production and consumption “under one roof”. Till now, it was not economically feasible, now it is, especially in areas with high electric rates.

Single-Family or Multi-Family Housing: The off-the-grid mode can readily be applied to Passivhaus-type, freestanding houses, or Passivhaus-type housing developments; the latter would have, say 16 pre-fabricated units to a building, 4 floors @ 4 units each, using centralized systems. The building would have a PV solar system on the roof, and/or have a parking area with charging stations for plug-in vehicles and a roof covered with PV solar panels. Energy use per household would be significantly less than for a Passivhaus-type, freestanding house.
The use of pre-fabricated units, built under modern, factory-controlled conditions, would ensure their quality and energy efficiency, a great improvement over stick-built in the field. The units could be of less-costly, standard design, or of more-costly, custom design. With the energy-efficient foundations and basements in place, the units would be erected into a weather-tight structure within about one week, ready for finish work.
Pre-fab, multi-family, Passivhaus-type, housing will become more prevalent going forward. It would greatly reduce the operating and maintenance costs and improve the livability of housing for at least 50% of US households. Customized pre-fab for single-family and multi-family housing is highly advanced in Europe, particularly in Denmark, Sweden, Finland and Norway. Vested interests have prevented its widespread implementation in the US for decades. Here is how this would work for a freestanding house.
Off-The-Grid: Flexibility is important for living off-the-grid. If one energy source is inadequate, another should be available to supplement. My starting point is a relatively NEW, freestanding house, similar to a Passivhaus, NOT grid-connected, with properly angled rooflines, proper solar orientation and passive solar features, and using about 70% to 80% less energy per square foot per year for heating, cooling, and electricity than a relatively NEW, standard, code-designed house. For living off the grid, in a near-zero-CO2-emission mode, the house would need to be equipped with:
– A 10 kW, roof-mounted, PV solar system + a battery system wired for 48 V output, with charge/discharge controller + an LP-fired, DHW heater with 200-gallon storage tank and DC electric heater to enable the use of PV solar energy + a system with DC pump and water-to-air heat exchanger.
– An LP-fired, 2 – 3 kW AC generator to periodically charge the batteries to about 90%, in case of insufficient PV solar energy during winter, due to fog, ice, snow, clouds, etc.

– A whole house duct system to supply warm, cool, humidity-controlled air, with an air-to-air heat exchanger to take in fresh, filtered air and exhaust stale air at a minimum of 0.5 air changes per hour, ACH, per HVAC code.

– For space cooling, a small capacity, high-efficiency AC unit would be required on only the warmest days, as the house would warm up very slowly.

– For space heating, 1 or 2 LP-fired, vented heaters, plus 2 or 3 V-120, vertical, LP storage tanks (each holds 96 gal); required on only the colder days. The tanks would also supply the LP-fired generator.

– A plug-in EV, such as a Nissan, or plug-in hybrid, such as a Chevy-Volt, would be charged with DC energy from the house batteries by bypassing the vehicle AC to DC converter (reduces inverter losses), provided the house batteries have adequate remaining storage energy, kWh, for other electricity usages.

Any excess electricity would bypass the already-full batteries and go to the DC heater in the DHW tank. Any excess thermal energy would be exhausted from the DHW tank to the outdoors.

DHW heating and space cooling would be mostly with the ample PV solar energy available during spring, summer and fall, thereby significantly reducing the LP consumption for DHW. See Note 8.

In winter, several days may pass with minimal PV solar energy. Electrical energy storage would be required in less sunny areas, such as New England and Germany. A 10 kW PV solar system would produce about 4.32 kWh on an overcast winter day in New England. See Note 4. This is insufficient, as the house may need about 10 kWh/d on overcast, winter days. See Note 8. The battery system and LP-fired engine-generator would be needed for 1 to 2 hours and the plug-in EV would need to use a public charger. The rest of the year, the PV solar system would have greater outputs.

Household Energy Management: To determine the capacity of the energy systems, list all the energy users on a spreadsheet, how much they use (amp-hours/day) and what time periods they are on and off. The sum will give the hour-to-hour energy consumption per day, or per week. Subtract the hour-to-hour PV energy generation to yield the hour-to-hour surplus (charges the batteries) or deficit (discharges the batteries). Energy consuming items can be scheduled on and off to manage the energy flows. If there is a prolonged period of no sun, the engine-generator and the batteries supplement any solar energy. Having as many DC devices as possible reduces DC to AC conversion losses.

Total Site Energy: Site energy, based on an assumed 3,000 miles of public charging of the plug-in, an assumed 150 hours of LP-fired generator operation, and an assumed no DHW heating with PV solar, would be 36,783,512 Btu/yr. On the grid, without PV solar systems, it would have been 51,645,516 Btu/yr. On the grid, with PV solar systems, it would have been 29,616,212 Btu/yr

Investments and Energy Cost Savings: An absorbed glass mat, AGM, battery system costs about $350/100 Ah. A 1,550 Ah system, wired for 48 V, sufficient for about 4 days, would cost about $10,000 installed. See Note 4. A PV solar system costs about $4,000/kW of panels. An 10 kW system would cost about $40,000 less subsidies.

On the grid, in a standard, code-designed house, no PV solar system, annual bills would be for electricity $1,080, space heating + DHW $2,100, and plug-in $648. Off the grid, in an energy-efficient house, the electric bills, including public charging of plug-in would be about $600, and for space heating + DHW + LP generator operation would be about $1,276, plus there would be the bonus of mostly free DHW heating and space cooling during spring, summer and fall, when ample PV solar energy would be available. See Note 8.


The above alternatives clearly show to provide off-the-grid, standard (mostly energy-hog) houses with PV solar systems, and electrical and thermal storage systems, they would need to be of such large capacity the costs would be prohibitive, if “net zero-energy and near-zero CO2 emissions” is the goal.

As a result of better building practices and materials much more energy-efficient houses can be constructed. Such houses, equipped with efficient mechanical and electrical systems, and the lower cost PV solar and battery systems, enable more and more homeowners to “live off the grid”, plus charge a plug-in vehicle.

PV systems have at least 25-year useful service lives, and battery systems, if property operated, have at least 10 to 15 year useful service lives. The homeowners will be enjoying annual cost savings for heating, cooling, electricity and gasoline for decades that are sure to increase, at about 2 to 4%/yr., year after year, plus they have the satisfaction of minimizing their CO2 emissions “footprint”.


NOTE 1: If an EV travels 12,000 m/yr. at 0.30 kWh/mile, 3,600 kWh/yr., or about 10 kWh/d, would be required, equivalent to the production of a 3 kW PV solar system in New England. Gasoline cost avoided = 12,000 mi/yr. x 1 gal/28 mi x $3.50/gal = $1,500/yr. 

NOTE 2: Because PV solar systems have become much less costly, it would be less complicated and lower in O&M costs to increase the capacity of the PV solar system to also provide electricity to heat DHW, thereby reducing the propane for the DHW heater, instead of having an $8,000 roof-mounted solar thermal system for DHW; no tube leaks, freeze-ups, less moving parts. With a properly insulated, large capacity DHW tank, say 250+ gallons, there would be enough DHW for 5 – 7 days.

NOTE 3: A maximum of about 50% of battery nameplate rating is available. To prolong the useful service life well beyond 8 years, say 12 – 15 years, batteries should typically be charged to a maximum of 95% and discharged to not less than 75%; shallow cycling. Very rarely should they be discharged to a minimum of 45%, as deep cycling reduces life. Also, life is prolonged if charging and especially discharging is slow; a few amps for many hours is much better than many amps for a few hours. Depth-of-Discharge, DOD, factor = 100/(95 – 45) = 2.0.


– Battery charging loss is about 10% and discharging loss is about 10%, i.e., input 100 kWh, store 90 kWh, output 81 kWh. 

– Inverter DC to AC efficiency, about 25% at 2% of rated input, is about 90% from 20% to 100% of rated input.

– Minimizing DC to AC conversion by using DC devices (fans, pumps, heaters, etc.) avoids battery and inverter losses.

House, Passivhaus level, Low, High and Average Energy Usage:

House low daytime energy draw from battery for 1 hour, inverter operating at 0.2/10 = 2% of capacity = 0.2 kW x 1 h x 1/0.25 inverter eff x 1/0.9 battery loss = 0.89 kWh, or (1000 x 0.89) Wh/12 V = 74 Ah; PV solar contributes

House high daytime energy draw from battery for 1 hour, inverter operating at 2/10 = 20% of capacity = 2.0 kW x 1 h x 1/0.9 x 1/0.9 = 2.47 kWh, or 206 Ah; PV solar contributes.

House average energy draw from battery over 24 hours, inverter operating at less than 20% of capacity = 2855 kWh/yr kWh/yr/365 d = 7.821 kWh, daily consumption x 1/0.8 x 1/0.9 = 10.862 kWh, or 905 Ah; PV solar contributes.

PV Solar Winter, Summer and Average Energy Generation:

PV solar energy to battery; overcast winter day = 10 kW x 3 h x 0.16 CF x 0.9 battery loss = 4,320 Wh; battery system and generator would be needed. 

PV solar energy to battery; sunny summer day = 10 kW x 6 h x 0.70 CF x 0.9 battery loss = 37,800 Wh; excess energy may be used to charge plug-in, heat sauna, hot tub.

PV solar daily energy to battery averaged over one year = 10 kW x 24 x 0.14 CF x 0.9 battery loss = 3,024 Wh

Example of determining required battery capacity: 

Provided to house……………………………………..10.00 kWh/d AC

Provided by 10 kW PV system………………………4.00 kWh/d AC

Provided by 3 kW generator………………………….3.00 kWh/d AC

From DC to AC inverter……………………………….3000 Wh/d AC

Inverter loss factor……………………………………….0.90

To inverter………………………………………………….3333 Wh/d DC

Wiring loss factor………………………………………….0.90

From battery………………………………………………3704 Wh/d DC

Battery discharge loss factor…………………………….0.90

From battery adj’d for discharge loss…………………4115 Wh/d DC

Autonomy period…………………………………………….4 days

From battery during autonomy period………….16461 Wh DC

Depth of Discharge factor……………………………..0.30; a low value for longer life

Charge in Battery………………………………………54870 Wh DC

Temperature loss factor………………………………..0.90

Charge in battery adj’d for temperature…………60966 Wh DC

System voltage……………………………………………..48 V

Battery system capacity………………………………1270 Ah

Battery aging factor………………………………………0.85

Battery system capacity adj’d for aging…………..1494 Ah

Battery system to have 2 strings in parallel; each string with 12 batteries in series

Rating of selected battery……………………………..750 Ah

Battery strings in parallel………………………………….2; 3 strings is acceptable, if necessary

Battery system rating…………………………………..1500 Ah

Battery voltage………………………………………………..4 V

Batteries in series…………………………………………..12

Total number of batteries………………………………….24

Battery cost = 1,500 Ah x $350/100 Ah = $5,250, plus the cost of wiring, charge/discharge controller, 48 V to 120 AC inverter, mounting racks, and installation, for an installed total cost of about $10,000.

Such a 4-day event may occur only a few times during winter. At other times, PV solar generation would be greater and the battery discharge % would be less, which reduces battery aging. Energy generation would be sufficient for DHW heating (supplementing the LP heater of the DHW system) and for most of the plug-in vehicle charging.

NOTE 4a: If a household had a 10 kW solar system, and a highly efficient, NOT off-the-grid house, plus a plug-in EV, then the DC from the PV system could charge a 7 kWh TESLA wall-hung battery unit and the DC output of the battery would charge the car battery. In that case there would be no 10% – 15% inverter loss.

There would still be the normal car battery charging loss, which could be 4% or greater, in addition to the 8% “round-trip” DC loss of the TESLA unit. However, that unit would not be sufficient to charge an EV, so at least 2 units would be required. A household with 2 EVs would need at least 3 – 4 units.

NOTE 4b: If a household had a 10 kW solar system, and a highly efficient, off-the-grid house, plus a plug-in EV, the above system of batteries would charge the EV during the year, except during most winter hours, plus serve the whole house.

NOTE 5: As space heating and cooling would be required for a small percentage of the annual hours, an air-source heat pump would be overkill and too expensive in this case.

NOTE 6: Because inverters have lower efficiencies at PV solar outputs of less than 20% of inverter capacity (occurring mostly during winter, and dawn and dusk throughout the year), the monthly energy feed-in ratio is about 4/1 in New England. In Southern Germany, further away from the equator, it is about 6/1. See monthly output from 2 monitored solar systems in Munich. In Vermont, the hours of sunshine ratio is about 2.54 and production ratio is about 3.8 for fixed-axis systems. Here are two field-mounted examples, one fixed-axis, one 2-axis tracking.

The Ferrisburgh Vermont solar farm, 1,000 kW, south-facing, correctly angled, field-mounted, has monthly averages of 4 years of production that show the monthly energy feed-in ratio of July/December = 1.000/0.263 = 3.80, and a 4-yr average CF = 1,323,879 kWh/yr/(8,760 hr/yr x 1,000 kW) = 0.151.

The South Burlington Vermont solar farm, 2,200 kW, 2-axis tracking units, field-mounted, has monthly averages of 4 years of production that show the monthly energy feed-in ratio of July/December = 4.936, worse than fixed-angle, and a 4-year average CF = 0.167, which is .167/.151 = 10.6% better than fixed-angle, even though such trackers are claimed to be up to 45% better! In Vermont, the better performance of 2-axis, up to 21%, occurs mostly during May, June, July and August. Snow would readily slide off the panels at the steep winter angles. Such systems would be about 25% to 30% more costly and require greater O&M expenses, which will reduce any economic advantage.

NOTE 7: Whereas, the daily or weekly maximum solar output of Germany may be up to 60% of installed capacity, kW, during a very sunny period, it may be near zero, due to fog, ice, snow, clouds, etc., during winter. As a result, Germany’s mix of PV solar systems (old and new, dusty or not, partially-shaded or not, snow/ice-covered or not, fog/cloud-shrouded or not, facing true south or not, correctly-angled or not) has a low nationwide capacity factor of about 0.10. This compares with a New England CF of about 0.12; the theoretical CFs are about 0.12 for Germany, about 0.143 for New England.

NOTE 8: The Passivhaus specifications for all climates are:

– Space heating* ……. 15 kWh/m2/yr or less………15 kWh/m2/yr x 3413 Btu/kWh x 1 m2/10.76 ft2 = 4,758 Btu/ft2/yr

– Space cooling ………. 15 kWh/(m2a) + 0.3 W/(m2aK) x Dry Degree Hours

– Domestic hot water…. 12 – 35 kWh/m2/yr; depending on number of occupants

– Primary energy+…….120 kWh/m2/yr or less…. 120 kWh/m2/yr x 3413 Btu/kWh x 1 m2/10.76 ft2 = 38,074 Btu/ft2/yr

– Airtightness………….. 0.6 ACH or less @50 Pascal

The above implies the following values:

– Space heating demand^…. 10 W/m2 or less…………10 W/m2 x 3.413 (Btu/hr)/1 W x 1 m2/10.76 ft2 = 3.2 (Btu/hr)/ft2

– Space cooling demand……..10 W/m2 or less…………10 W/m2 x 3.413 (Btu/hr)/1 W x 1 m2/10.76 ft2 = 3.2 (Btu/hr)/ft2

– Total energy to site………….42 kWh/m2/yr or less….42 kWh/m2/yr x 3413 Btu/kWh x 1 m2/10.76 ft2 = 13,300 Btu/ft2/yr

– Window U-values……………0.8 W/m2.K or less…….0.8 W/m2 x 3.413 (Btu/hr)/1 W x 1 m2/10.76 ft2 x 1 K/1.8 F = 0.141 (Btu/hr)/ft2/F; equivalent to R-7.1

– Door U-values………………..0.8 W/m2.K or less……..0.141 (Btu/hr)/ft2/F, equivalent to R-7.1

– Air-to-air heat exchange eff…….80% or greater

* Applies to space cooling in warm climates.

+ The 120 kWh/m2/yr is primary energy for space heating and cooling, domestic hot water, auxiliary electricity, domestic and common area electricity. 

^ A 2,000 sq ft (186 m2) Passivhaus would need for space heating a 10 x 186 = 1,860 W, say 2 kW, thermostat-controlled, electric heater in the fresh air supply duct.

NOTE 9: The space heating consumption would be partially met by indoor heat sources, such as lights, cooking, computer, refrigerator, passive solar heat gains, and people, with the rest by the LP-fired space heaters; this would reduce LP consumption!! A Passivhaus-style house could have dark, stone floors that would warm up due to passive solar heat gains and would give up the heat in the evening when the curtains would be closed.


IECC Energy Codes: The 2012 International Energy Conservation Code will require more insulation, a tighter envelope, tighter ducts, better windows, and more efficient lighting than the 2009 code.

Blower-door testing became mandatory: The 2009 infiltration threshold of 7 ACH @ 50 Pascal became 5 ACH @ 50 Pascal for climate zones 1 and 2. It became 3 ACH @ 50 Pascal for climate zones 3 through 8. All homes in zones 3 through 8, and some homes in zones 1 and 2, will be required to have a whole-house mechanical ventilation system. Almost all houses built before 2009 have greater ACH values, i.e., they are energy hogs during summer in warm climates (such as the US Southwest) and during winter in cold climates (such as the US Northeast).

Passivhaus Standards: Energy-efficient, Passivhaus-level, construction requires at least 0.6 ACH @ 50 Pascal, R-40 walls, R-60 roof, R-20 basement, 85% efficient air-to-air heat exchanger, R-7 windows and doors, high-efficiency appliances and lighting.

An R-20 basement can be achieved by using 4 inches of 100 psi, Dow blue board (two sheets of 2’ x 8’ x 2” thick, special order at Home Depot) UNDER the 18” wide concrete footing, and 4 inches of standard, 25 psi, Dow blue board UNDER the basement slab and on the OUTSIDE of the concrete walls, and continued, at 4″ thickness, up the OUTSIDE of the house walls to the roof eave. All seams, below and above ground, must be staggered and taped; exposed blue board must be covered with stucco, or with ½” PT plywood that is solid-color stained to match concrete.

Thus the concrete basement serves as a thermal mass that cools slowly in winter and warms slowly in summer, which reduces indoor temperature variations and the annual energy for heating, cooling and electricity. See Note 8.

NOTE: The full basement of an average 2,000 sq. ft. house, requires about 80 cu. yd. of concrete, which causes about 50,400 pounds of CO2 emissions. Adding up to 40%, by weight, flyash to the mix will significantly reduce CO2 emissions. 


Moisture inside walls is one of the biggest challenges of energy-efficient home building in colder climates. Such moisture will cause wood rot, mold and odor, which starts at a wall internal relative humidity of about 80%. Indoor moisture may be due to cooking, bathing, plants and many other sources. Outdoor moisture may be due to precipitation that leaks past the siding or due to soil moisture that is drawn into below-grade assemblies.

The Residential Exterior Membrane Outside Insulation Technique, REMOTE, manual describes in detail how to build energy-efficient building envelopes and minimize moisture issues in colder climates. The manual contains numerous drawings and images. See Figure 1.

Standard Stud Wall, Foam Outside of Sheathing, Fiberglass Wool in Wall Cavities: In climate zones 5, 6, 7, 8, and 9, at least 2 – 6 inch (R-10 – R-30) of rigid foam board, such as Dow blue board (extruded polystyrene, XPS, R-5/inch, Perm 1, vapor retarder), all seams staggered and taped, needs to be applied to the OUTSIDE of the ½” plywood sheathing of 2 x 4 and 2 x 6 walls; OSB is not recommended for sheathing, as it is more prone to rot. Filling the cavities of a 2 x 4 wall with fiberglass wool will add a nominal R-13, of a 2 x 6 wall will add a nominal R-19. To avoid condensation inside the stud walls, interior/exterior ratios of R-values apply, based on climate zone. See page 28 of above “REMOTE” URL and below Example.

NOTE: Since 1987 the National Roofing Contractor’s Association (NRCA) has recommended designers use R-5.6/inch as a reasonable estimate of the actual thermal performance of polyisocyanurate insulation over the lifespan of a wall or roof assembly, because R-values are less at lower temperatures. NRCA average R-value for aged polyiso is 5.7/inch at 75 F; some manufacturers claim 6.2/inch. See URLs.

– R-values of fiberglass, EPS and XPS become higher at lower temperatures.

– R-values of Polyiso, including ZIP System sheathing, become lower at lower temperatures.

– Most R-value testing is at 75 F, with 50 F on the cold side and 100 F on the warm side.

– XPS should be used in colder climates.

Example: In climate zone 6 (Vermont), to satisfy REMOTE requirements:

For a nominal R-28 wall 3.5” of fiberglass wool (R-13) and 3” of XPS (R-15) is needed

For a nominal R-33 wall 3.5” of fiberglass wool (R-13) and 4” of XPS (R-20) is needed

For a nominal R-38 wall 3.5” of fiberglass wool (R-13) and 5” of XPS (R-25) is needed

For a nominal R-44 wall 5.5” of fiberglass wool (R-19) and 5“ of XPS (R-25) is needed

Indoor relative humidity conditions are ‘low’ (20% RH in winter), ‘normal’ (30% RH in winter) and ‘high’ (40% RH in winter). The worst case, 40% RH, with a dew point of 40.1 F, was chosen for below calculations.

The inside of sheathing temps, (T indoor – Delta T x R cavity/R total), are:

For the R-28 wall: 42.7 F at 17 F, 39.5 F at 10 F, 34.8 F at 0 F; two temps are below the dew point

For the R-33 wall: 46.1 F at 17 F, 43.3 F at 10 F, 39.4 F at 0 F; all but one above the dew point.

For the R-38 wall: 48.6 F at 17 F, 46.2 F at 10 F, 42.8 F at 0 F; all well above the dew point.

For the R-44 wall: 44.3 F at 17 F, 41.3 F at 10 F, 36.9 F at 0 F; all but one above the dew point.

Standard 2 x 6 Wall, Foam Outside of Sheathing, 3″ Foam in Wall Cavities: A wall could have 4” XPS on the outside of the sheathing and 3” XPS, snugly fit, in the wall cavities, for a nominal R-35, and no fiberglass wool, leaving a 2.5” space (R-0.77) for wiring and piping. At outdoor temps of 17 F, 10 F, 0 F, the inside of sheathing temps would be 44.4 F, 41.4 F, 37.1 F, respectively, all but one above the dew point.

– Water piping and electrical wiring can be run inside the exterior stud walls. The 2 x 6 studs can have 1″ x 1″ cutouts for 3/4″ copper tubing.

– Heat from the warm piping in the wall will increase the temperature of the cavities and drywall.

– The interior surface of the wall will “feel” warm/more comfortable, as the cold temperatures are mainly in the exterior foam.

– The exterior foam board will greatly reduce air infiltration and minimize any condensation at the exterior of the sheathing. A moisture barrier, such as Tyvek DrainWrap or 6 mil of polyethylene, must be applied to the exterior of the sheathing.

– The traditional 6 mil polyethylene vapor barrier on the interior of the sheetrock is not needed and must NOT be applied to allow the wall to dry to the interior of the house.

– The stud wall, including sheathing, will be much warmer than a standard stud wall without external foam, and will attract much less water vapor on colder days, and will need much less “drying out” time on warmer days than dense-packed cellulose in 12” thick double walls. See next section. The foam in the stud wall will need minimal “drying out” time, as it absorbs almost no water vapor.

Standard 2 x 4 Wall, Cellulose in Wall Cavities and on Exterior: The humidity within various wall designs was measured for a year. The results indicate a 2 x 4 wall with 3.5” of cellulose (R-13) in the cavities and latex paint vapor retarder, with 4.3“ of cellulose (R-16) on the exterior of the sheathing, would result in internal RHs of up to 80% for only a few hours of the year. The exterior of the cellulose layer would have house wrap on the inside and outside. Adding more external cellulose would reduce internal RHs, as the cavity wall would be warmer. This wall has the disadvantage of not having 2.5” for wiring and piping, as would the above-mentioned Alternative Example wall.

Double 2 x 4 Walls With Cellulose In Cavity: Building a 12” wall with an outer 2 x 4 wall, plus an inner 2 x 4 wall, studs staggered, creates a thermal break. The entire space, filled with densely packed cellulose (R-3.5/inch), creates 3 heat flow paths:

R-path 1 = 1.5″ wide with R-3.5 for outer stud + {(12 – 3.5) x 3.5 R/inch = 29.75 for cellulose} = 33.25.

R-path 2 = (8 – 1.5) = 6.5″ wide with 12 x 3.5 R/inch = 42.

R-path 3 = same as R-path 1.

However, on cold days, the indoor moisture gets drawn towards the inside of the sheathing. It condenses or freezes on the cellulose close to the sheathing and on the sheathing, which is the coldest surface, because inside the wall, the temperature is below the dew point or below freezing! Dense-packed cellulose may become, by weight, up to 25% water!

The dew point would be 12 – (65 – 40.1)/{(65 – 17))/12} = 5.78 inch from the sheathing wall at 17 F outdoor temp, and 6.57 inch at 10 F outdoor temp.

– The freezing points would be 3.75 inch, and 4.80 inch, respectively, from the sheathing wall.

– The conditions would be worse for Paths 1 and 3, as they have lesser R-values.

– The R-values of damp and frozen cellulose are less than of dry cellulose.

– Thicker walls with densely packed cellulose are slow to dry on warmer days.

– Water piping cannot be run inside the exterior stud walls, as it might freeze.

– The interior surface of the wall will not “feel” warm/more comfortable, as the cold temperatures are mainly in the cellulose.

– The cellulose, mainly recycled newspaper, absorbs water, would mold, unless treated with borate to make is mold resistant, but the sheathing, a structural component, if not enough dried on warmer days, will mold and rot.

Conclusion: XPS foam insulation on the OUTSIDE of the sheathing combined with fiberglass wool or XPS foam in the stud wall cavities, as shown in the above example, is the superior, long-term approach for high R-value walls.

NOTE: A thin layer of spray foam applied to the interior of the stud wall sheathing and the sill area “to reduce air infiltration” on new houses, with the rest of the stud wall cavity filled with fiberglass or cellulose, must allow the wall and sill area to dry to the interior of the house, i.e., no polyethylene moisture barrier applied at the inside of the dry wall, plus no moisture barrier applied to the outside of the sheathing and sill area, to allow the sheathing and sill area to dry towards the exterior. On retrofit jobs, with existing moisture barriers, these requirements would be costly to implement.

NOTE: Attics usually have very high humidity, up to 90%, as water vapor is lighter than air and gable vents often do not provide enough ventilation on windless days. A layer of spray foam applied to the rafters and roof sheathing must allow the roof sheathing to dry to the exterior of the house, i.e., no tarpaper, ice and water shield, or other moisture barrier must be applied under the shingles!

Photo Credit: Housing and the Grid/shutterstock

Joris van Dorp's picture
Joris van Dorp on Nov 26, 2014

Willem, interesting article and reasoning, thank you, especially for providing your calculations which helps a lot.

I’m curious about your battery system cost estimate.

You say 100 kWh/24V AGM battery system cost comes down to $8000. This implies a cost of $80 per kWh total system cost.

I found the following supplier of 12V AGM batteries, but they sell at about $500 per kWh, excluding value-added tax, and that is just for the batteries excluding the balance-of-system costs for a working system in the home. These batteries weigh just under 30 kg per kWh, so a total system would weigh about 3 tons, presumably. Balance-of-system costs for a three ton battery system would not be zero, I suspect. This makes me suspect that such a system would cost far more than $8000.

Calculation for 12V/26Ah AGM battery sold by the above shop, priced at €124 a piece:

€124 / (12V * 26 Ah/1000) * $/€ = $500/kWh. (ex.VAT)

Could you provide more information on how a 100 kWh home battery system will be about $8000 inc. VAT rather than at least $50000 ex.VAT as per my calculation above? Or is the sales price I’m using much higher than asked by other suppliers of AGM batteries?

Thanks in advance 

Another thing I am wondering about is the issue of energy taxes. In the Netherlands, households pay a lot of tax on electricity, in fact it’s mostly tax what they pay, the cost of the actual electricity being a minor part of the electric bill. This is fine of course, because it reduces the need to tax households via other routes for government revenue, and because it stimulates households not to waste too much electricity. However, owners of solar panels supplying 100% of their own electricity are exempt from paying this tax, so the financial motive of ‘going solar’ in the Netherlands is completely explained by the interest in not paying tax (and additionally perhaps, the interest in wasting lots of tax-free electricity?). Indeed, without this energy tax exemption, ‘going solar’ is not financially viable in the Netherlands and would only be done by hard-core solar enthousiast home-owners with money to burn. The situation in Germany is similar I suppose. What do you think of this situation in relation to the question of the economics of going solar, on-grid or off-grid?

Jeffrey Miller's picture
Jeffrey Miller on Nov 26, 2014


This site lists eleven 12V batteries with ah > 110. Prices are about half of the UK price you got:

> price <- c(363,427,549,349, 538,549, 643,629,435, 651,645 )

> ah <- c(125,150,200,200, 210, 210, 230,245,250, 255,255)

> ppkwh <- round( 1000* price/(ah*12) )

> summary(ppkwh)

   Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 

  145.0   212.0   214.0   209.1   231.0   242.0  


The two least expensive batteries ($145 per kwh) are from the same manufacturer and are on sale. I have no idea whether the batteries differ in quality, but a price of ~ $200/kwh seems ball park right, before tax. In the US, sales taxes vary by state and county from 0% to around 10%, depending on your state. (I’m not sure what number Willem is using for $/kwh). 


Edit: It looks like Willem is assuming $200 for a 12V 100ah system, which works out to $166 per kwh, which is a bit low but doable from the above figures. 

Edit 2: Looking more carefully, Willem seems to sometimes assume a 12V system and sometimes a 24V system. He seems also to use the 12V prices for the 24V at fixed ah which is not right. So I agree that his prices seem off by a factor of two. 

Jeffrey Miller's picture
Jeffrey Miller on Nov 26, 2014


I think the point of contention at least as far as battery prices go is your $8,000 estimate. From note 4, it seems like you have in mind a 100 kwh system. The price for such a system would be around $17,000 using your value of $200/100ah at 12V.  While it is possible that battery prices could decrease by a factor of two if more people were to buy them, it is by no means guaranteed. 

The larger point, which Joris alludes to and which is the topic of this post in a slightly different context is that it can’t be socially optimal for us all to be buying batteries and sticking them in our basements or sticking solar panels on our roofs. People have a romantic attachment to the idea of distributed power, but it must be the case that the total cost of distributed power systems greatly exceeds the cost of an integrated system, because you eliminate the advantages to scale and even more importantly you eliminate the advantages of averaging  or optimizing (optimizing across the whole system will yield a much lower cost than optimizing each little piece separately and then adding them together). So in this sense subsidies which encourage us to all go out and build little solar arrays and buy batteries, especially in climates where we don’t get much sun, are socially harmful. 

I like your ideas on stricter building codes and moving toward passive house designs  and we should probably aim toward this ideal going forward, but the fact is that our enormous existing building stock is not going away quickly and there is only so much that one can do to retrofit an old building (25% energy savings is ambitious). 

John Miller's picture
John Miller on Nov 27, 2014

Willem, I too have been studying and analyzing distributed power and energy systems over the years.  A few issues caught my attention in reviewing your Post that you might want to consider addressing in the future:

·         The efficiency of solar PV low voltage DC to 110V AC power ‘inverters’ are normally in the 80-90% range.  Generally only industrial scale inverters have higher efficiencies compared to average inverters installed with Residential solar PV systems.  This factor also decreases the capacity factor of a given solar PV panel array system (max. design generation).

·         The same 80-90% efficiency range generally applies to 110V AC to 12-24V DC power ‘converters’ for most EV chargers.  I am sure you are aware the most efficient systems are 220-440V AC EV chargers; which also minimizes charging cycle times.  Also, the somewhat popular/perceived technology of using EV batteries to supply the centralized Power Grids repeats these energy conversion loss levels twice; which makes the practicality of this technology approach to backup power for variable wind power highly questionable.

·         Most ‘off-the-grid’ Residences that require or desire reasonable reliable power 24-7-365 also install backup generators; petroleum or natural gas motor fuels.  Even a multi-ton Residential battery system will be inadequate in different geographic locations and times of the year.  Based on personal experience, Bellingham, Washington was subjected to 90 consecutive (cloudy) days without the sun one year.  Without backup generators, you will be in the dark much of the time in many locations around the world.

·         The most efficient Residence, besides insulation, sealing, orientation to the sun, etc. also requires major behavior adjustments to not over heat the house during the winter, over cool it during the summer, efficiently manage external-internal air exchange, not waste hot water, etc.  Most of these systems can be automated including external air exchange, but the critical variable is how the Resident programs the system to minimize energy waste.  This generally means wearing a sweater in the house during the winter and shorts/’T’ shirt during the summer.  And, of course, as Mom use to tell us as kids, “turn out the lights” (when not needed).  Even this function can be automated (motion-timer switches more commonly found in some Commercial Buildings).

Jeffrey Miller's picture
Jeffrey Miller on Nov 27, 2014


Your new price estimate of $6000 for a 100 kwh battery system looks even less plausible to me than your old estimate of $8000. Where can you buy these kinds of batteries at that price, $60 per kwh? Do you have a source for that? 

It seems to me that you are assuming in your calculations that the price per amp hour is roughly constant, regardless of voltage, which doesn’t seem right. I don’t see any 48V batteries for sale anywhere (maybe I am missing something – let me know if you have a source), so to get a 48V potential it would seem to me that you have to wire four 12V batteries (or eight 6V batteries) together in series. So yes, the required amp hours goes down by a factor of 4 when you go from a 12V to a 48V system, but the number of batteries required goes up by exactly the same amount – I don’t see how the economics change at all by changing the voltage. What basically determines the cost of your system is the number of kwhs of energy you want to store, not the voltage you want to maintain. 

On distributed vs non-distributed, I believe the only way to we can quickly and with certainty drastically reduce our carbon emissions and at the same time provide plentiful, cheap, on demand energy is by building large numbers of new nuclear plants (CCS will also need to play a big role) and this needs obviously to be centralized. We should be pooling our collective resources to build as many new Gen III nuclear plants as we can, as efficiently, safely, and quickly as we can, and not waste time and scarce economic resources on what can only ever be partial solutions like wind and solar, whether distributed or otherwise. (I might feel differently about this if wind and solar were not intermittent or if there were some reasonable certainty of cheap, scalable, and environmentally friendly storage; but these sources are intermittent and I am not at all assured that we will ever have cheap storage).




Nathan Wilson's picture
Nathan Wilson on Nov 27, 2014

As much as we’d all like to have an off-grid mansion on our own private island (with a solar heated fresh water swiming pool), this lifestyle is much more energy intensive than living in cities.  Not just for heating and lighting, but especially personal transportation and transport of goods and services.

In every developing country, every farmer knows that some (most?) of his children will grow up and move to the city, not just for jobs, running water, and electricity, but for entertainment and shopping too.

We would never expect that apartment dwellers could save money by going off-grid and putting a solar array on their patio.  It is obviously much cheaper to wire the whole building, and use a common power source, no matter the type.  The same applies to dense campuses of building, and likely urban neighborhoods (rural electrification via grids has always been expensive per household though).

When you say that an off-grid home will use a small gasoline generator, that brings to mind the portable lawn-mower-derived unit that are advertised for $400 every storm season.  These are only designed to run a few hours a time, and are not suitable for off-grid use.

Here is a vendor that sells propane/ng powered standby generators, starting at $2100 for a 6 kW unit.  To that, add $340 for a transfer switch, and $90 for a starter battery.  You also need to pour a concrete pad on which to mount it, plus installation.  

You have not demonstrated the conditions underwhich your off-grid home is more economical than grid power (again I suspect it is only for rural locations).  Clearly, within apartment buildings, it’s cheaper to combine households into a “centralized” system.  When grids are being built from scratch (i.e. developing nations), I suspect the economies of scale combined with the benefit of letting the power company pay the up-front cost of the centralized system mean that in cities in general, centralized is cheaper and better for the poor and businesses alike.

It’s a cheap shot to complain about 80% of energy being wasted; you have not demonstrated a more efficient solution.  Nor have you demonstrated that your technology does not become more affordable when centralized.  This source reports that only 6% of electricity is wasted as losses in the transmission and distribution of US electricity; that is much less than the efficiency loss of switching from a 60% efficient NG combined cycle plant to a distributed sub-10kWatt piston generator (35% at best?).  It’s much easier to regulate pollution from the centralized generator as well.

Centralization is important for nuclear (for obvious reasons) but renewables as well.  With solar or wind, centralization and grid aggregation smooths the effect of passing weather, likely for lower cost and lower environmental impact than using larger batteries.  Centralization reduces maintenaince cost, keeping old equipment in service longer (less e-waste to be recycled).  Also, centralization allows higher performing resources from distant locations to be used.  This means that lower use of fossil backup will result.  Power-to-fuel technology may be miniaturized for residential use someday, but for today it works much better on a MWatt scale.

Bill Hannahan's picture
Bill Hannahan on Nov 28, 2014

Willem, I have copied the last part of your calc with corrections in paren ().


  • 11111 Wh/d/48 DC system volts = 231 Ah/d. (at 48 volts)
  • 231 Ah/d x 1.11 battery temperature derate factor x 6 days autonomy x 1.4 DOD factor = 2,153 Ah. (at 48 volts)
  • 2153 Ah (at 48 volts)/100 Ah individual  ( 48 volt) battery capacity  = 22 (48volt) batteries , round off to 24 (48volt batteries);  batteries with greater Ah may be used.
  • 48V system voltage/12V battery voltage = 4 (12 volt) batteries in series (per 48 volt battery)
  • (24 48 volt batteries) x 4 (12 volt batteries per 48 volt battery)  = (96 12 volt) batteries.
  • System cost (96 12 volt batteries) x $300/100 Ah = ($28,600)


At 15 cents per kWh the house uses $548 worth per year, about 2% of the battery cost alone. Add up the cost of maintenance, solar panels, batteries, inverter, charger, and depreciation. Run a long term lifecycle calculation using a reasonable interest rate and battery life.

Add in the risk of having two tons of lead in every house, the risk of an acid spill, fire, explosion, electrocution.

It just makes no sense, EVEN IF SOLAR PANELS ARE FREE, unless you live in the boondocks.



Joris van Dorp's picture
Joris van Dorp on Nov 28, 2014

I think this is getting confusing now.

Identical battery cells wired in series will give a voltage that is the sum of the voltages of the individual cells, but the total capacity (Ah) of this arrangement will remain the same as the capacity of a single cell.

Conversely, connecting battery cells in parallel will increase the capacity (Ah), but not the voltage.

What this boils down to is that the power storage capacity (Wh) of a battery is equal to the sum of the storage capacity of the individual battery cells, whether you connect them in series or in parallel. Connecting them in series or in parallel (or a combination) lets you choose the required voltage, but does not let you increase the power storage capacity (Wh). There is no free lunch, as it were.

So your battery setup of 4 by 5 cells of 100 Ah/12V each will give a total battery with a voltage of 48V and a capacity (Ah) of 500 Ah, giving a power storage capacity of only 22.5 kWh. So you would need four of such batteries to reach 90 kWh of power storage capacity.

Bill Hannahan's picture
Bill Hannahan on Nov 28, 2014

Willam, a few thoughts;

1)   Regarding the modified note 4. The assumption that the solar array will meet 46% of demand on an overcast winter day contradicts the more realistic Note 7.

2)   You have a section called “Investment and Energy Cost Savings:” Yet you never calculate a dollar amount for off grid savings. I think it would be a negative number.

3)  You have a section called “Energy Efficient House, Off the Grid, With PV Solar System and Plug-in Vehicle“, Yet you do not have a section called “Energy Efficient House, ON the Grid, With Plug-in Vehicle”

Comparing an energy hog on the grid with an efficient home off the grid is not fair. At 15 cents per kWh the house uses $548 worth per year. Add up the cost of maintenance, solar panels, batteries, inverter, charger, and depreciation. Run a long term lifecycle calculation using a reasonable interest rate and battery life.

Add in the risk of having two tons of lead in every house, the risk of an acid spill, fire, explosion, electrocution.

It just makes no sense unless you live in the boondocks.

4)   Our home electric bill covers about 1/3 of all the electricity that supports our lives. Where should the other 2/3 come from?



Bill Hannahan's picture
Bill Hannahan on Nov 29, 2014

Willem, Thanks for the thoughtful comments.

1)   “the “Off-the Grid” alternative is “near-zero CO2 emission””. Somewhat true in operation, but there are a lot of emissions in the manufacturing of all that equipment.

Averaging the numbers over the whole year does not tell the whole story. The car will consume a lot more energy / mile in winter, due to heating load, higher drag from slush and snow, and reduced battery performance, just when the solar output is at its minimum.

How Technology Could Reduce the Cold Weather Drain on EV Batteries | MIT Technology Review

The cold winter water takes a lot more energy to heat than tepid summer water. I think the generator will run more than you expect in winter.

2)   “You mention the 3653 is 1/3 of your electricity.”

I was referring to the fact that our electricity consumption is divided into roughly three equal parts, home, commercial and industrial. we pay for one directly, the others in the price of goods services and taxes we pay. Going off the grid reduces 1/3 of the energy we consume.

Joris van Dorp's picture
Joris van Dorp on Dec 1, 2014

I see you have corrected the article partly by proposing a new system for the off-grid passiv house: one which relies more on the diesel generator + includes the contribution from solar under overcast conditions, I think.

But will the average household energy demand be the same during an overcast winter week s it is durign the year? The passiv house will not have stored passive heat if there is little or no sun, and there will also be a greater demand for lighting during such days. So a passiv house will presumably be using most of its energy during exactly such periods of the year. It will not be an ‘average’ weekly energy demand profile. I guesstimate that your energy demand assumption of the passiv house for an overcast winter week is about 50% too low at least. You will need a bigger battery or a bigger/longer-running generator, I suspect.

Nathan Wilson's picture
Nathan Wilson on Dec 2, 2014

“…charging the plug-in in winter may not be feasible; public charging stations would need to be used.”

BEV in the future will likely come in two battery sizes, 60-120 mile range and 150-300 mile range.  The longer range cars will be compatible with public charging.  

The minimum range BEVS will mostly be charge daily, and will work better when plugged-in all night in the winter, so the batteries can be kept warm, which improves the range.  Tesla claims the cold-weather degradation is small, but then again, the Tesla batteries are big to begin with.

Jeffrey Miller's picture
Jeffrey Miller on Dec 2, 2014

I noticed a signifcant drop off in range in my Tesla last winter, 20 to 30% or more depending on temperature. This winter so far it was shaping up to be a similar sized effect. I had my car in for service yesterday and they noticed that the louver which controls airflow to the battery wasn’t working properly and fixed it. It think it might have been stuck open. This morning when I drove to work the temperature was 24 degrees F. Normally at that temperature my wh/m would have been around 400 on my 13 mile commute. This morning it was well under 300, which is typical for summer driving (in summer I sometimes get ~260). I am going to monitor this going forward, but it seems possible that my poor mileage last winter was caused at least in part by the stuck louver keeping the battery too cold.

I agree with Willem’s comments about the Volt. When I had my Volt, my summer range was around 48 miles, but in the winter that dropped to the high 20s.


Willem Post's picture
Thank Willem for the Post!
Energy Central contributors share their experience and insights for the benefit of other Members (like you). Please show them your appreciation by leaving a comment, 'liking' this post, or following this Member.
More posts from this member

Get Published - Build a Following

The Energy Central Power Industry Network is based on one core idea - power industry professionals helping each other and advancing the industry by sharing and learning from each other.

If you have an experience or insight to share or have learned something from a conference or seminar, your peers and colleagues on Energy Central want to hear about it. It's also easy to share a link to an article you've liked or an industry resource that you think would be helpful.

                 Learn more about posting on Energy Central »