Chris Williams Solar Hero Style

In this article, I’ll go through the basic step-by-step process of how to evaluate, understand and communicate the financial benefits of investing in a solar thermal system. The analysis will be on the client side, but obviously it’s critical for sales as well.

Before you read: get familiar with financial terms and analysis, you should read the first article in the series “Finance 101 for Renewable Energy Pros”. Also, it’s important to note that I’m using the word “finance” as a way to build financial models, understand the economic drivers and benefits of specific  technology – not finance as in ‘we financed our car instead of paying cash’.

Here are the other articles in this series:

  1.  Finance 101 for RE Pros
  2. Finance 101 for PV Pros
  3. Finance 101 for Geo Pros

We’ll be going through the same drill that I did with solar PV and geothermal in terms of the outline but the specific content will be tailored to the technology that we’re looking at, solar thermal.

Here’s the outline

  1. What makes SHW special and a little different then analyzing other technologies
  2. Step 1. Estimating solar thermal load, array size and power production
  3. Step 2. Gross and net installed costs
  4. Step 3. Determine the value of a SHW BTU
  5. Step 4. Estimating operations and maintenance costs
  6. Step 5. A few examples IRRs and sensitivity analysis for residential and commercial projects based on 1) load 2) fuel source 3) site characteristics
  7. Marketing Implications
  8. What I did not address that could be investigated.

A few issues around the difficulties and issues with determining the exact NPV of a SHW system. 

  • On residential applications, it’s too costly to figure out exactly how much hot water is being used. Thus, we use assumptions that frankly, are not very accurate. See the Canadian study that found out the average of 65 gallons used per day, was actually around 44.
  • Unless the hot water generator is the only fuel source of that specific kind, it’s difficult to estimate on residential applications and mainly based on assumptions, which can be very wrong.
  • On commercial applications, it is common to use ultrasonic BTU meters for a week or so to understand exactly how much water is being used. However, it’s still key to understand daily and yearly usage patterns. For example, if a laundromat is used heavily in the morning or a college dormitory is not used during the summer that will have implications for the value of the heat the solar thermal system is creating. See point 2.
  • Production and usage of solar thermal energy are not equal. A property owner only gets the value of a solar BTU when they’re using water that is getting preheated by a solar thermal system. If they’re not using water, and the solar thermal system is producing that energy gets lost. Not all of it is lost, because the storage tank is able to hold a lot of water but they can’t hold it forever. The reason this is important for financial modeling is because, UNLIKE SOLAR PV, just because the solar thermal modules produce power doesn’t mean it was used and thus doesn’t mean the financial benefit was realized. The classic example is a family that goes on vacation for 2 weeks, if it’s a pressurized solar thermal system (we’re not going to get into pressurized vs drawback in this article and the design and financial implications of each) the pump will likely still cycle and energy will be produced, but nothing will be used. From a finance perspective, nothing is gained, only lost in the power the pump needed to run.
  • Quoted prices for solar thermal systems can vary widely from site to site and between geographic regions. The main drivers between sites will likely be 1) structural support needed. All else equal pitched shingle roofs are cheaper then flat roofs. 2) If a storage tank is required. For buildings that have a constant load 365, storage is typically not required. Pool heating is a good example. This will decrease installed costs. Between geographic regions that main drivers of costs tend to be the training of the crew. Almost all of the parts are off the shelf, or close to it, so it’s difficult to get better pricing on equipment, however a crew’s ability to executive and their level of training will be different between regions.
  • Module output is based on more factors then in solar PV. In a solar PV product output is mainly based on 1) the solar resource available 2) orientation of the module 3) efficiency of the module 4) temperature. With solar thermal, all of those factors also apply IN ADDITION to the load profile of the building. Why? The higher the load of the building the colder the water will tend to be, all else equal, when entering the solar therm module. This will increase heat exchange. So for example, if the modules were 180 degrees, the water passing through them will collect more BTUs if it enter the modules at 50, then if it entered the modules at 100. What this means is that if we installed 10 modules on a building with a load of X, if the same number of modules were installed on a building and the load was 2X, the production of the modules would be much higher. For this reason, it’s a good idea to keep the solar fraction low in a design, to maximize the BTU production of each module. How low? Dr. Ben suggestions between around 30% and 60%, see his great explanation of the subject here.
  • Maintenance costs can vary widely based on the type of system, equipment used, equipment warranties, and what the type of system is connected to. Also, because the solar thermal industry is relatively small, I haven’t been able to find large data sets of warranty information that I can be confident in.

 So, let’s walk through some examples. 

The main financial driver of solar thermal is —-> Number of modules * Net Cost to install each module / Value of Power Produced by the array. O+M is also a factor and we’ll discuss this in a section below.

Two Examples where we will perform many types of sensitivity analysis based on fuel type, roof type, and load to see how this will impact installed cost, and production.

Each will be located in Boston.


  • Fuel Types: Oil, electric, natural gas
  • Number of Residents: 2, 4, 6


  • Load: Laundromat
  • Roof Type: Flat vs pitched shingle
  • Fuel Type: Oil and NG

Step 1. Determine the number of modules to install and estimating power production

The purpose of this post is not to get into an in-depth engineering discussion but to highlight how engineering decisions will impact the financials of a project.

The first step is to determine the load of the building, and then determine the number of modules and how many BTUs each module produces. It’s key to remember the low of dimension returns and to not oversize a solar thermal system. A single module will produce the most BTUs per module, but the least overall. Between a 30% and 60% solar fraction tends to have the best output based on installed costs.

LOW solar fraction will produce more BTUs because production is largely depend on the delta between EWT and the module, higher delta = more heat exchange and more power. You can also manipulate the production of each module with the same solar fraction by changing the size of the storage tank, which will also have diminishing returns because larger tanks cost more. All else equal, smaller tanks will lead to less production, because the water will heat up fast, and larger tanks will produce more given the same load and array size.

Good rules of thumb that we’ll use in this article:

Massachusetts: 100 therms per year per module for 20% to 40% solar fraction on commercial. Residential 50 to 70 therms per module as solar fraction nears 80%.

This is for a standard flat plat module, not EVs and yes, it will vary somewhat based on the module.

Note: We would need to adjust these rules of thumb based on location. Greater insolation = more output, all else equal and assuming the rest of the array characteristics are the same.

Based on the load, solar resource, and available roof space, determine how many modules the system will be. In most commercial applications this will be handled by an engineer on the project.

Again, keep in mind that when determining the number of module to use, as the solar fraction increases (the amount of BTUs used that are produced by solar) each additional module will create more BTUs for the system, but each module will produce fewer BTUs, so it comes down to a question of costs per module vs. production per module.

Here’s a good rule of thumb to calculate the expected therms needed to heat hot water based on average daily hot water use.

Here’s how we do it.

  •  A – Guess the number of gallons used per day using ASHRAE calculations for residential or commercial applications
  • B – Calculate the BTUs needed to heat that water to the site set point, normally 120, sometimes higher in commercial applications.
  • C – Calculate the BTUs in a year and divide by 100,000 to get therms.
  • D – This will give you the rough therms needed per year. On residential applications it’s best not to go above an 80% solar fraction, on commercial we keep it much lower.

For example,

 Residential Base Example

  •  1 – A 4 family household uses 65 gallons per day (20 for the first person, 15 for each additional.
  • 2 – They require 37,901 BTUs each day to heat water –> 65 gallons X 8.33 (the specific heat of water, BTUs needed to heat 1 gallon 1 degree F) X 70 degrees (this is the expected delta between the street water and 120, assuming the street water is 50 degrees. It will be hot in the summer, cool in the winter)
  • 3 – 37,901 X 365 = 13,834,047 BTUs per year, or 138 Therms.
  • 2 Modules @ 50 therms per module. On residential projects, where the solar fractions tends to be high, we typically assume a module will produce between 50 and 70 therms depending on the load. In this case, we would select 2 modules and expect they would produce about 100 therms per year.
  • MAKE SURE TO USE MODELING SOFTWARE For This. The above information is based on my experience and recollection doing residential site visits and designs. After you’ve done a few you’ll also start to use these rules of thumb, but they should be used for real client proposals, I’m using them in this article because they can still communicate the points I’m trying to address.

Commercial Base Example

Let’s assume the next project is a laundromat that uses 1000 gallons per day. If we follow the same above example, we’ll say that it needs 2128 therms per year.

30% of 2128 is 638 therms.

So, let’s assume that we install 8 modules and each produced around 100 therms per year, equaling 800 therms.

Step 2. Determined installed costs Gross and Net

After we have an estimate of modules numbers, we need to determine how much they will cost to install.

This is the investment for the client and it’s also heavily impacted by federal, state and utility incentives.

Below are the line items that will drive the gross installed costs.

 Gross Installed Costs Based on 

  • Equipment. Modules, Pipe, Pumps, Storage Tank, etc.
  • Direct Labor – Depends on your specific margin and if it’s prevailing wage. It’s typical for roofers or carpenters to do the roofing work and a plumber to do the utility room work and potentially running pipe on a flat roof.
  • Permitting – Depends on the AHJ
  • Roof Warranty – On commercial projects, getting a letter from Firestone, etc will often cost a few thousand off the bat.
  • Structural Specs – You may or may not have to have a PE do some structural drawings for you. If you do, it’ll cost you.
  • EPC Margin – Typically between 20 and 30%.

 Government Incentives

 Government Incentives will vary widely based on

  • 1) Commercial vs Residential
  • 2) Specific state
  • 3) Type of client (building or fuel type)
  • 4) Specific utility or municipal

Right now, I’ll just run through some of the best incentives, but obviously each state will be different.


  • 30% ITC – Self explanatory for most in the renewable energy industry.
  • MACRS – Only for Commercial Clients . See the solar PV example for how to calculate MACRS.

State Incentives 

State incentives get pretty diverse, below I’ll explain a few more of the major ones. They will differ by state, city, and utility. It’s important to understand exactly how the incentive is calculated and the maximum incentives.

Massachusetts – Massachusetts has several incentives available.

  • MA CEC Commonwelath Solar Grant is $25 X SRCC C Rating for the Module X Number of Modules. Different Max Incentives for residential and commercial.
  • State Tax Credit – Residential – 15% with a max tax credit of $1,000
  • Cambridge Energy Alliance – $2,000 cash rebate up to 50% of the cost of the system. MUST BE LOCATED in Cambridge. All types of customers.
  • City of Boston – Renew Boston Solar Program: $3,500 cash rebate for residential SHW. MUST Be located within the city of Boston. Only available to residential
  • So for example, a residential customer in Boston would be eligible for MA CEC Commonwealth Solar Grant, State tax Credit and City of Boston rebate in addition to the federal ITC.
  • I wanted to show the two local incentives just to show that there are numerous incentives and you need to be aware of them.
  • NOTE: The excel model I’ve created was only created with Massachusetts in mind

New York

New York has interesting legislation around SHW that is offsetting electric.

  • NY Personal Tax Credit: Residential and Multi-Family. 25%, MAX $5,000
  • NYSERDA Solar Thermal Incentives. This is an interesting incentive, it can apply to residential and non-residential. Here is the wording: “$1.50 per kWh displaced annually, for displacement of up to 80% of calculated existing thermal load”. Installation requirements: “System must generally supplement an existing electric water heater; system losses due to shading and orientation may not exceed 25% of ideal production without losses”. What this means is that this rebate is only good for SHW that are directly replacing electric water heating.
  • The cash rebate is based on the estimated electric displacement in year 1. So for example, for a residential home, if the system was expected to generate 100 therms per year (this number would actually need to be generated by an approved software provider for the NYSERDA grant) , and it was offsetting electric, we would need to convert this to kWh to calculate the rebate amount.
  • Here is how we would do this:
  • There are 3,412 BTUs in 1 kWh. Or, if you have therms, the number is .03412 (because there are 100,000 BTUs in a therm). 1 therm / .03421 will result in kWh
  • So 100 therms / .03412 = 2,930 kWh that the system will displace.
  • 2,930 X $1.50 = $4,396 cash rebate for a residential system that displaces electric.
  • NOTE: The excel model I’ve created was only created with Massachusetts in mind. You can use it for New York, but you’ll to change how the rebate structure is valued.

Washington DC

Washington DC program converts therms into MWhs and then those MWh can be sold into the REC market. This is for every type of customer and fuel source. They key to remember about DC program is that it’s performance based, which means a payment is received whenever SRECs are minted. There are some reporting requirements depending on the size of the SHW system.

  • Currently, SRECs are selling in DC for around $300 / MWh:
  • So, the calculation is similar to what it was for NY.
  • NOTE: The excel model I’ve created was only created with Massachusetts in mind, in order to use this for Washington DC, you’ll need to change how the model is calculating the incentives.


  • So 100 therms / .03412 = 2,930 kWh = 2.9MWh, equivalent that that the system will create.
  • Only full MWh’s can be applied for SRECs so 2 SREC x $300 = $600 for the yearly production.


  •  For our commercial example that produces 800 therms per year, here is the calculation and numbers.
  • 800 therms / .03412 = 23, 446 kWh or 23 MWh, which would be worth 23 X $6,900 at today’s SREC prices.

California Solar Thermal Incentives

California is offering significant cash rebates for SHW systems that are directly replacing NG.

Here’s basically how it works, it’s based on estimated NG therm (for low-income) and NG and electric for the general program displacement in year 1, based on software outputs.

Normal Program:

Amount: Step 1 Incentive Rates (contact utility to determine current incentive levels): Systems that displace natural gas: $12.82 per estimated therm displaced. Systems that displace electricity: $0.37 per estimated kWh displaced
Maximum Incentive: Step 1 Incentive Limits (contact utility to determine current incentive limits): Single-family residential systems that displace natural gas: $1,875. Single-family residential systems that displace electricity: $1,250.Commercial and multifamily residential systems that displace natural gas: $500,000Commercial and multifamily residential systems that displace electricity: $250,000

Low-Income Program:

Amount: Step 1 Incentive Rates (contact utility to determine current incentive levels): Single-Family Low-Income: $25.64 per therm displaced. Multi-Family Low-Income: $19.23 per therm displaced.
Maximum Incentive: Single-Family Low-Income: $3,750. Multi-Family Low-Income: $500,000

For our previous commercial example, if these 8 modules were on a low-income housing facility in California, the rebate would be

800 X $19.23 = $15,384

To note this difference keep in mind that the same production would only qualify for a $4,600 incentive in Massachusetts.

If it was not a low-income facility, the value would be $12.82 per therm OR

800 X $12.82 = $10,256, still 100% higher then MA incentives.

**Note: A system that produces 800 therms in Boston, will produce between 20% to 30% more in Southern CA because insolation is so much higher. So 800 in Boston, might be around 1,000 in Southern, CA.

Step 3. Determine the Value of the Value of a Solar Thermal BTU.  

Once we have determined a) the size of the array b) the net installed cost and c) the estimated power production, we need to understand the VALUE of the power production. The value a solar BTU is worth exactly the same as a BTU that it is replacing. Thus, we need to understand how to calculate and understand fossil fuel BTU costs. In general, this is the equation.

BTU Cost = Price Per Unit of Energy / System Efficiency.

Depending on the fuel source, we’ll apply a correction factor to the above number to convert it back into therms.

Value of a Traditional BTU

  • NG Example: $1.2 per delivered therm / 95% efficient boiler = $1.26 per delivered therm.
  • Oil Example: $4 per gallon / 90% efficient boiler = $4.44 per delivered gallon. However, there are 140,000 BTUs in one gallon of fuel, so we need to correct it to therms, which have 100,000 BTUs. 100,000 / 140,000 = .714. This is our correction factor. $4.44 * .714 = $3.17 is the cost of a delivered “OIL THERM”.
  • Electric: For electricity the calculation can seem a little more confusing, because electric is considered to be 100% efficient, so we don’t need to account for the efficiency of the system, but we need to confirm kWh (the cost of the fuel) into therms. kWh to BTU conversion is 3,412, so kWh to therms is .03412 –> $.15kWh / .03412 = $4.39 per electric therm.

In the excel spreadsheet, the calculations are done for you on the “Fuel Cost” table.

 A few examples, 

So, just to be clear from our residential example. If the system produces 100 therms, the value of the those 100 therms is almost 100% determined by the fuel is offsetting.

Here’s what the 100 therms is worth (given all my above assumptions about fuel costs and efficiency are true)

VS NG = 100 X $1.26 = $126

VS Oil = 100 X $3.17 = $317

VS Electric = 100 X $4.39 = $439

 Step 4.  Estimating maintenance costs and lifetime costs. 

There is a lot that goes into O+M and lifetime costs of the system. For my model, I will use .5% of installed costs per year for O+M costs. In truth, I haven’t been able to find a good data set that I can trust for SHW systems and thus determine the value more on a case by case basis.

The O+M costs will largely be determined by

  • 1 – System type. drain-back tend to be cheaper then pressurized
  • 2 – Equipment warranties on modules, tanks, pumps
  • 3 – Site specific conditions: located next to salt water, etc

Step 5 Example of IRR’s based on project specifics

Now, I’m just going to walk through a few examples to see how some variables will change the specifics of the project.

We’re going to take our 2 base residential and commercial examples from above perform some sensitivity analysis on 1) installed costs 2) fuel type 3) load

All the systems will be based in the greater Boston area, but not directly in Boston or Cambridge.

Residential Base Case – 2 Modules, 100 therms produced per year, 40 Gallons of hot water usage per day

A – Fuel Source: How Does the IRR of the project change based on the fuel source? Let’s assume inflation will be at 4% for each fuel. 

Oil IRR = 5.71%

Why so low? With only 2 people living in the home for the base case, the therms produced by each module will be low, thus it won’t offset a large amount of Oil.

Electric: 11.93%.

Why? Electricity is a really expensive BTU.

Natural Gas: -2.79%.

A NG therm is only worth $1.29 with today’s prices, so it’s hard to make this sell.

B ) Load -What happens if we increase the load? When changing the load, we’ll stay with oil is the fuel source.

4 people – 65 Gallons – 65 Therms Produced Per Module. IRR = 9.09%

6 people – 95 – 75 therms per module = 11.13%

C) Local Incentives. What happens if you’re client can apply for an additional city or municipal incentive?

Living in Cambridge with a $2,000 rebate. Oil, 2 people home. IRR = 15.58%. This shows that very specific and niche incentives can have a huge impact to a small amount of customers.

Commercial Base Case – 8 modules, 800 therms produced per year, 1,000 gallons of demand per day, NG fuel source. 

A) Pitched Roof vs Flat Roof

Pitched roof is slightly cheaper due to the less racking and potentially less structural work.  We’ll assume $4,100 per module (32 square feet) gross.

Pitched Roof IRR = 6.79%

A flat roof will require more racking and structuring engineering costs. Let’s assume it will cost ~ $4,600 per module.

Flat Roof IRR = 5.21%

B) How will fuel costs impact the IRR of commercial projects. Let’s assume we go with a pitched roof installed cost of ~4,100.

NG IRR = 6.79%, same as above

Oil IRR = 17.77%. Why so much higher than Residential? Two main reasons. First, each module will produce many more BTUs because the load is higher. Second, the system can receive MACRS depreciation.

Electric IRR = 26.94%

Implications for Marketing

IRRs are impacted by small differences in installed costs due 1) the site and 2) geographic-specific incentives and  this should have a large impact on your marketing activities. As you can see, IRRs can swing substantially, from negative 2%, to positive 27% with only a small change in the number of variables. This means crafting very specific marketing messages and targeting the right customers is very important for not wasting your time on selling SHW.

All else equal, leads from Cambridge or Boston are much more valuable then other leads from other locations with fewer incentives.

Also, clearly the fuel that it’s offsetting will have a large impact. A pitched roof located in Boston offsetting an electric fuel source with 6 people in the home is the absolute best case for SHW on a residential basis.

What I did not consider that a further analysis would go into:

  • Variable vs constant loads. I did not address the impact that variable loads can have on the value of the thermal BTUs being created. Fuel prices, especially oil and NG, fluctuate depending on the time of year. So, the time of year when the hot water will be used impacts the financial returns of a project. Constant vs variable loads will also have a large impact on how the system is sized.
  • Evacuated Tubes. Within the article, I didn’t discuss EV Tubes at all and simply used flat plates. I just did this to make the analysis simpler and easier and because I’m a huge fan of drain-back systems, which cannot be used with EV.  Here’s an article that evaluates flat plate vs. evacuated tubes.
  • Space heating. I did not go into examples for space heating in this examples. Although it’s technically possible, I’ve noticed that using it for space heating represents a small niche, within a small niche, so I didn’t feel it would be worth the time of going into the subject. If anyone is interested in this subject, let me know.