Community solar vs rooftop solar: which saves more over 20 years?

A subscription sells a discount on credits; a roof sells the kilowatt-hours themselves

The Department of Energy's community solar basics page describes the arrangement in two moves. Subscribers sign up for, or occasionally buy, a share of the output of a solar array built somewhere else, usually in the same utility territory, and then receive a credit on their electric bill for the electricity their share produced. You never own equipment, nothing touches your roof, and your utility still delivers your power and sends your bill.

The money works through two lines. The utility puts a bill credit on your account for your share of the array's output. The company that owns the project then charges you a subscription fee for that share, priced below the credit. Your saving is the gap between the two, and in programs that set the fee as a percentage, the gap is that percentage of the credit. New Jersey's official program page for subscribers says it plainly: you get a fixed percentage discount on the bill credits as your savings, and the remainder is paid as the subscription amount to the project owner. The discount applies to the solar credits, not to your entire bill, which still carries taxes, fees and fixed charges.

Rooftop ownership sells something different. You pay for the system, own every kilowatt-hour it produces, and keep the full value of each one under your tariff, minus the cost of keeping the equipment running. So the two options split the same electricity in opposite ways. A subscriber keeps a slice of the value with no capital at risk. An owner keeps all of the value after putting roughly $23,600 at risk on the benchmark system. Every number later on this page is a consequence of that split.

There is also a reason the comparison only applies to some readers. The same DOE page cites a national laboratory analysis finding that nearly 50 percent of households and businesses cannot host a rooftop system, because they do not own the building, the roof is shaded or too small, or the upfront cost is out of reach. For those households the question is not which saves more but whether a subscription exists where they live. DOE counts at least one community solar project in 44 states and localities including the District of Columbia, 24 of which have passed enabling legislation.

What five state programs actually promise subscribers, as of October 2026

The size of the discount decides most of the comparison, and it is the figure most often quoted without a source. The table below prints only what each state's agency, utility commission, legislature or official program administrator says on its own site, checked October 30, 2026. Where a program sets no minimum for ordinary households, the cell says so rather than filling in a market estimate.

Three patterns stand out. First, New Jersey is the only program in the table that sets a savings floor for every subscriber, and the March 2026 order applies the 20 and 25 percent minimums to newly registered projects. The state's subscriber page says discount rates can run from 10 percent to more than 40 percent, so a subscriber on an older project may hold different terms. Read the discount on your own contract rather than assuming the current minimum. Second, every other floor in the table is reserved for income-qualified households: Maryland's 10 percent (stated in the Commission's pilot-era FAQ, which the permanent program's regulations may revise), Oregon's 20 percent fee reduction and Colorado's 25 percent statutory discount do not bind offers to anyone else. Third, the one general-market figure an official site prints is an assumption, not a promise: Oregon's savings estimator assumes 5 percent net savings for a general market subscription and says those savings are not guaranteed except for qualifying low-income customers.

Several large community solar states are not in the table yet. New York, Massachusetts and Minnesota run long-established programs; they are omitted until each rule is confirmed at its official source, because this page does not cite third-party summaries of state rules. If you live in one of those states, the questions in the contract section near the end of this page apply in the same way.

The table also explains why the arithmetic below runs four discount levels rather than one. Five percent is Oregon's general-market planning assumption. Ten percent is Maryland's income-qualified floor. Twenty percent is New Jersey's floor for all new subscribers and DOE's best-practice target. Twenty-five percent is New Jersey's floor for low- and moderate-income subscribers and Colorado's statutory floor for income-qualified ones. Nothing in the comparison uses an unsourced industry range.

The inputs behind the 20-year comparison

Both sides of the comparison are computed by one short script, so every result can be checked. The rooftop side is the national benchmark system. The subscription side is sized to deliver the same credited energy, so the two options are buying the same kilowatt-hours and only the money flows differ. Every input that is not a published figure is labelled as an illustration, and you should replace those with your own numbers.

The value of a kilowatt-hour is the most important input and the one this page deliberately does not set. Electricity prices and their trends are covered by our sister site, and EnergyBS tracks how residential electricity rates are moving in 2026; your own bill gives the number that matters. We run 20 cents and 15 cents per kilowatt-hour as round illustrations, at annual escalation of 0, 2 and 3 percent, because the answer turns on exactly those two variables.

Two simplifications are worth stating before the results. The O&M figure is a national modelled cost, levelized so that every year carries the same amount; a real owner pays little in most years and a larger amount when a component such as an inverter fails, which the inverter replacement guide prices on its own. And the subscription is assumed to keep its discount for all 20 years, which is generous where contracts are shorter, carry escalating fees, or let the discount change at renewal.

Twenty years, nine price paths: who comes out ahead

The table sets the rooftop owner's 20-year net position, which is the value of everything the roof produced minus the $23,600 purchase and twenty years of O&M, against the total a subscriber saves at three of the program discount levels. The last column is the break-even discount: the subscription discount at which the two options tie over 20 years. A subscription offering more than that discount beats the roof on these inputs; one offering less loses to it. A negative break-even means the roof does not repay its cost within 20 years at all, so any subscription wins.

Read down the 20 cent rows first. With flat prices, a subscription at New Jersey's 20 percent floor beats ownership by about $1,570 over 20 years; with prices rising 2 or 3 percent a year, ownership pulls ahead by $4,200 to $7,700. The break-even column shows why: rising prices push the discount a subscription would need to keep pace from about 15 percent to about 37 percent, above every floor printed in the state table, short only of the 50 percent Colorado's statute sets for some federally supported income-qualified projects. At 5 percent, Oregon's general-market planning assumption, the subscription saves $1,717 to $2,296 at 20 cents, well below the roof in every rising-price case.

Then read the 15 and 12 cent rows. When each kilowatt-hour is worth less, the roof's fixed $23,600 has less to pay back, and at 12 cents it does not repay within 20 years in any scenario here, while every subscription still saves something because it carries no capital. That is the core asymmetry: a subscription's saving is a percentage, so it shrinks gracefully when credits are cheap, while ownership is a fixed bet that pays only above a threshold. The threshold is worth computing directly, and the section after next does that.

One household, two paths, followed to year twenty

Take the 20 cent, 3 percent case and follow both households year by year. The subscriber saves $360 in year one at a 20 percent discount and keeps saving a little more each year as the credit value rises, reaching about $1,892 in total by year five, $4,030 by year ten, $6,448 by year fifteen and $9,182 by year twenty.

The owner starts $22,072 behind after the first year's production and upkeep, still $15,501 behind at year five and $6,168 behind at year ten. The roof repays its cost in year 13 and stands $4,562 ahead by year fifteen, but that is still below the subscriber's $6,448. Only in year 17 does the owner's running total pass the subscriber's, and by year twenty it leads $16,871 to $9,182.

So on these running totals, before any resale value, the household that sells within 16 years, or whose roof needs replacing before then, would have been better off subscribing even on the case most favourable to ownership in this table. What an owned system adds to a sale price is a separate question, read in the home value guide. Under flat prices the crossover moves to year 22, beyond the 20-year horizon. The roof's advantage is concentrated in the final third of its life, which is why time in the house matters as much as the price of electricity.

Four separate reasons the gap moves

It is tempting to say ownership wins because you keep all the value, but that collapses several effects into one. Splitting them shows which inputs to check first on your own numbers.

The first effect is the share of value each option keeps. A 20 percent subscriber keeps one fifth of the credits' value; the owner keeps all of it. In the 20 cents, 3 percent case the twenty years of energy are worth $45,911, so the owner's extra share is 80 percent of that, $36,729. This effect always favours ownership and grows with every cent of rate increase.

The second effect is capital. The owner pays $23,600 in year one; the subscriber pays nothing up front. In nominal dollars that comes straight off the owner's side. The third effect is upkeep: the DOE O&M figure takes another $5,440 over 20 years from the owner, which the subscriber never pays because the project owner maintains the array. Subtract both from the first effect and you get the $7,689 gap in the owner's favour in that case. Run the same split at flat prices and the owner's extra share is only $27,472, which does not cover the $29,040 of capital and upkeep, so the subscriber wins by $1,568.

The fourth effect sits outside the 20-year window. The roof keeps producing after year 20, and the community subscription ends when its contract does. DOE's Federal Energy Management Program puts PV system lifespans at 20 to 30 years; on our inputs, years 21 to 25 add about $6,700 of net value to the roof at flat 20 cent prices, or about $14,094 if prices kept rising 3 percent a year. That tail is real, but it depends on the equipment lasting and on you still owning the house, so this page reports it separately rather than folding it into the 20-year totals. How long panels last in practice is the right check before counting on it.

The break-even credit value: a one-line rule

A simpler way to compare is to price the roof's electricity per kilowatt-hour. Over 20 years the benchmark system produces about 171,700 kWh on our inputs and costs $29,040 including upkeep, which is 16.9 cents per kWh in undiscounted dollars. A subscriber at discount d pays (1 minus d) times the credit value for each credited kilowatt-hour. Ownership is cheaper only when the credit value is high enough that the subscriber's price exceeds the roof's 16.9 cents.

Solve that and you get the rule. At flat prices, a 10 percent subscription beats the roof unless each credited kilowatt-hour is worth more than about 18.8 cents; a 20 percent subscription unless it is worth more than 21.1 cents; a 25 percent subscription unless it is worth more than 22.6 cents. If your bill's effective value per kilowatt-hour sits below the line for the discount on offer, the subscription saves more over 20 years; above it, the roof does.

Two adjustments make the rule honest. Rising prices lower the bar for the roof, because its cost is fixed while its value grows, so treat the threshold as an upper bound if you expect increases. And your own roof's price and production replace the benchmark: a quote at $3.40 per watt or a shaded roof producing 7,500 kWh pushes the roof's cost per kilowatt-hour up and the threshold with it. The cost-per-watt worksheet gives you the first number, and a production estimate for your roof gives you the second.

When your cash has another use

The totals so far are nominal: a dollar in year 18 counts the same as a dollar today. That flatters ownership, whose costs come first and whose gains come late. If the $23,600 could otherwise earn 4 percent a year, the comparison changes sharply, and it is worth seeing by how much.

Discounting every year's cash flow at 4 percent, the 20 cent, 3 percent case leaves the roof with a present value of $2,922 against $6,044 for a 20 percent subscription, and $7,555 at 25 percent. At flat 20 cent prices the roof's present value is negative, minus $3,811, while the 20 percent subscription is worth $4,697. At 15 cents the roof is underwater in both price paths. On these inputs, a household that treats its savings as having a 4 percent alternative use comes out ahead with a subscription in every case shown, provided it can get one at a meaningful discount.

Whether 4 percent is the right rate for you is a personal question, and many owners are not comparing against an investment but against a loan. Financing the roof turns the upfront cost into a payment stream with its own interest and fees, which the loan versus cash guide works through. The direction does not change: anything that makes money now worth more than money later tilts the comparison toward the option with no capital.

How your tariff tilts each side

Both options depend on what a credited or self-generated kilowatt-hour is worth, and the two values are not always the same. On the rooftop side, the value depends on whether your utility pays retail value for exported power. Under retail net metering, a surplus kilowatt-hour sent to the grid offsets one bought later, close to the 20 cent case. Under net billing, midday exports earn a lower rate, so the blended value of the roof's output falls toward the 15 or 12 cent rows. The net metering guide shows how to read that from your tariff, and time-of-use pricing adds a second layer, because the hour of production then sets the value.

On the subscription side, the credit's value is set by the program and the utility tariff, not by your contract. Illinois Shines tells readers that the credit for their portion of production is applied through net metering and the subscription fee is charged on top, and New Jersey's subscriber page describes the credits as the value of the electricity generated on the subscriber's behalf. Neither tells you the credit's cents per kilowatt-hour; your utility's community solar tariff or a recent subscriber bill does. If the credit is worth less than your full retail rate, the subscriber's saving shrinks in proportion, because the discount is a share of the credit.

The point for the comparison is to use the right number on each side. A household under net billing whose community credit is valued near retail can see the subscription look better than this page's matched tables suggest. A household under full net metering whose community credit is valued below retail sees the reverse. Rate structures and where they are heading belong to EnergyBS, so check the specific rates there and on your bill before plugging them into the rule above.

Roof, tenure and moving: the filters that come before the math

For many readers the arithmetic is moot because one option is not available, and a landlord faces a different filter again: the roof is available but the bill may belong to the tenant, which the rental property solar guide prices as a break-even rent premium. Renters and most condominium owners cannot install a system on a roof they do not control, and an association can limit or decide placement on common property; the HOA solar rules guide sets out what five state access laws allow. A roof that needs replacing within a few years makes a rooftop installation more expensive than the benchmark, because the panels have to come off and go back on, and the roof-first guide prices that sequence. Heavy shade lowers production and raises the cost per kilowatt-hour above the break-even line, as the shaded-roof guide measures.

Expected time in the home is the filter most often skipped. The example above shows the roof overtaking a 20 percent subscription only in year 17 on the case most favourable to ownership. An owned system can add value when you sell, but that value is not guaranteed to equal what the system has not yet earned. A subscription, by contrast, moves with you inside the same utility territory in New Jersey and Maryland according to their official pages, and in New Jersey it can be cancelled with no fee. If you move to a different utility, you generally have to cancel and re-enroll with another project.

Income is the third filter. Every guaranteed floor in the state table except New Jersey's general one is reserved for income-qualified households, and some programs reserve capacity for them as well: New Jersey's March 2026 announcement says at least 51 percent of program capacity is reserved for low- and moderate-income subscribers, and Maryland requires projects to serve at least 40 percent of their output to them unless subscribers own the system outright. The income-qualified programs guide covers what rooftop help still exists for owners who qualify.

Reading a subscription contract before you sign

The arithmetic assumes a clean contract: a fixed discount on the credit, no fees, credits sized to your use, and freedom to leave. Real offers vary on every one of those points, and the official pages in the state table are specific about where they vary.

Start with how the fee is set. A percentage-off-credit contract tracks the credit's value, so your saving rises and falls with it. A fixed per-kilowatt-hour price, or a fee that escalates each year, can do worse: Oregon's program FAQ notes that some subscriptions include a fee that increases each year and that pay-as-you-go plans charge a per-kWh price that rises over time. If the fee escalates faster than the credit's value, the saving shrinks and can turn negative, the same escalator problem that drives the lease and PPA comparison.

Then check the size. Credits beyond what you use do not always carry full value. Maryland pays out excess credited energy at a rate that may be below the subscription cost and caps combined subscriptions at 200 percent of baseline use; Oregon suggests aiming for no more than 90 percent of typical annual use; Illinois warns that oversized subscriptions can leave you paying for credits you do not use. Finally, read the exit terms. DOE's best practice is no exit, termination or sign-up fees; New Jersey's program states none, while Oregon tells general-market subscribers to understand the early-cancellation process and any fees before signing a contract of at least 10 years. Illinois Shines adds one more distinction worth knowing: a community solar subscription is not the same as a green or renewable supply offer from a retail electricity supplier.

Contract terms to pin down before subscribing

• Get the discount written as a percentage of the bill credit, and confirm whether it is fixed for the term or can change at renewal.
• Ask whether any fee escalates each year, and by how much, and compare that with how the credit's value is set.
• Find the credit's value per kilowatt-hour on your utility's community solar tariff or a subscriber bill.
• Size the subscription to your annual use, below the program's cap and ideally under it.
• Read the cancellation clause for fees, notice periods and what happens if you move outside the utility territory.
• Check whether you qualify for an income-qualified floor in your state, since most guaranteed minimums apply only there.
• Keep the program's disclosure form, and know where your state's consumer complaint process is.