What NFPA 110 Requires for Stored Generator Fuel

For hospitals and other facilities operating under the CMS adoption of NFPA 99-2012, the referenced edition is NFPA 110-2010. In that edition, the annual fuel quality requirement appears in Section 8.3.8. Later editions reorganized and revised some of these provisions, so always verify the edition adopted by your AHJ before citing section numbers.

Which NFPA 110 sections actually govern stored fuel?

Four sections carry the direct fuel obligations, and two more shape them. They sit in three different chapters of NFPA 110, Standard for Emergency and Standby Power Systems: Chapter 5 sizes the supply, Chapter 7 governs the installed fuel system, and Chapter 8 governs maintenance and testing.

Quoted language is from the 2019 edition. Chapter 7 sections as compiled by Curtis Power Solutions, NFPA 110 installation requirements; Chapter 8 sections per NFPA 110 maintenance and testing. See the edition section below before citing a number to a surveyor.

Read those together and the structure becomes clear. The code requires a large tank, requires that the fuel in it be clean when the engine calls for it, and gives you exactly two ways to make that true over time: burn the fuel before it degrades, or manage it.

Why 7.9.1.3 is the requirement most facilities fail

Section 7.9.1.3 gets at the problem that is easiest to overlook with standby fuel: diesel does not stay in the same condition indefinitely just because nobody is using it.

The section essentially gives you two paths. Size and operate the system so the fuel is consumed within its storage life, or have provisions for dealing with fuel that becomes stale or contaminated. In many standby applications, especially hospitals, data centers and other facilities carrying large emergency fuel inventories, normal generator exercising simply does not turn the fuel over very quickly.

Think about what happens in practice. The generator may run every month, but usually only for a relatively short exercise period. NFPA 110 has historically required at least a 30-minute monthly exercise, with diesel generators meeting specified loading or exhaust-temperature conditions. Even when you add periodic extended testing, that can still represent very little fuel consumption compared with a tank designed to support many hours—or even days—of emergency operation.

That creates an important distinction: exercising the generator is not the same thing as exercising the fuel inventory.

A generator can accumulate perfectly acceptable exercise records while much of the diesel in the main tank remains there year after year. How quickly that fuel deteriorates depends on its composition and the conditions in which it is stored. Temperature cycling, water intrusion or condensation, microbial contamination, oxidation and accumulated sediment can all shorten its useful storage life.

Water is especially important because it tends to collect at the lowest point in the tank. Once you have persistent water, you have created conditions that support microbial contamination at the fuel-water interface. The resulting biomass can contribute to sludge, filter plugging and corrosion while remaining largely invisible from the top of the tank.

That is why 7.9.1.3 matters so much. If normal consumption is not turning the inventory over fast enough to keep the fuel suitable for use, somebody has to actively manage the condition of that fuel. Waiting for the annual generator test to reveal the problem is not fuel management.

What counts as "a supply of clean fuel" under 7.9.1.2?

NFPA 110 requires the fuel system to provide clean fuel to the engine, but it does not give you one numerical definition of "clean." That is where laboratory testing, the applicable fuel specification, engine-manufacturer requirements and the condition of the storage system have to be considered together.

ASTM D975 provides the basic specification for diesel fuel and includes requirements for properties such as water and sediment, sulfur, cetane quality, distillation, flash point and corrosion. For No. 2-D S15 diesel, for example, the sulfur maximum is 15 ppm. ASTM D975 also establishes a minimum cetane number or calculated cetane index of 40 for the applicable grades.

Stored-fuel testing should go beyond simply asking whether the fuel resembles acceptable product at the point it was delivered. You also want to know what has happened to it while it has been sitting in the tank.

That is why we pay particular attention to water. ASTM D2709 measures water and sediment by centrifuge, while ASTM D6304 Karl Fischer testing can measure much smaller quantities of water. Bell commonly uses 200 ppm by Karl Fischer as an action point for stored standby diesel. That is a fuel-management threshold, not an ASTM D975 limit, but an upward water trend is something we want to investigate before it becomes free water at the bottom of the tank.

And where you take the sample matters.

A sample pulled from the middle or upper portion of a tank can tell you what the bulk fuel looks like while completely missing the water, sediment and microbial debris collecting underneath it. For a stored-fuel program, bottom sampling gives you information that a convenient sample from higher in the tank may never reveal.

"Clean fuel," then, should not be treated as the result of one number. It means having fuel suitable for the engine and a storage system that is not quietly accumulating the things most likely to keep that fuel from reaching the engine when it is needed.

What does the annual fuel quality test have to cover?

NFPA 110 tells you that fuel quality has to be tested at least annually using appropriate ASTM standards. What it does not do is hand you a universal laboratory panel and say, "Run these eight tests."

That distinction matters.

ASTM D975 is the basic specification for diesel fuel, and it identifies important fuel properties and the ASTM methods used to measure them. Those properties include things such as distillation, flash point, water and sediment, sulfur, corrosion and cetane quality. They are all useful when you are trying to determine whether stored diesel remains suitable for use.

But stored fuel creates additional questions that cannot be answered simply by treating the tank as though you were qualifying a fresh delivery of diesel.

How much water is accumulating? Is microbial activity developing? Are the results changing from one sampling event to the next? Is contamination concentrated at the bottom of the tank even though the bulk fuel still looks acceptable?

For standby-fuel applications, Bell's Mission Critical testing slate uses ASTM methods including D86 distillation, D93 flash point, D130 copper strip corrosion, D287 or D4052 for API gravity/density, D976 calculated cetane index, D2709 water and sediment, D6304 Karl Fischer water and D2622 sulfur. Together, those results give you a much better picture of what has happened to fuel during storage than any single test can provide.

Even that is still a snapshot.

Microbial contamination is a good example. A tank can develop a biological problem between annual laboratory tests, particularly when water is present. ATP-based testing using ASTM D7687 provides a way to monitor microbial activity more frequently and establish a trend instead of waiting for visible biomass or plugged filters to tell you there is a problem.

That is the difference between annual compliance testing and an actual fuel-management program. The annual test gives you a documented condition at a point in time. Periodic monitoring tells you which direction the tank is going.

Which NFPA 110 edition is your surveyor enforcing?

Before you put an NFPA 110 section number into a hospital policy, make sure you know which edition actually applies.

For facilities operating under the federal CMS Life Safety Code framework, CMS adopted the 2012 editions of NFPA 101 and NFPA 99. NFPA 99-2012, in turn, references the 2010 edition of NFPA 110. That is why a hospital can be inspected today against requirements from an NFPA 110 edition published more than fifteen years ago.

That matters because section numbers and, in some cases, the requirements themselves have changed between editions.

In NFPA 110-2010, the annual fuel quality test appears in Section 8.3.8. Later editions reorganized Chapter 8 and moved the annual fuel-quality provision. More recent editions have also continued to revise the language surrounding fuel testing and fuel-system requirements.

The same caution applies to fuel capacity. The familiar 133% tank-capacity requirement appears in older editions of NFPA 110, including the edition incorporated through the CMS health-care code chain. More recent editions have revised the fuel-capacity provisions, particularly for longer-duration classifications.

There can also be additional state, local or facility-specific requirements layered on top of the CMS requirements.

So don't assume that the newest NFPA 110 on the website—or the edition somebody used at another facility—is automatically the one governing your system. Identify the adopted edition and the AHJ first. Then write the policy around the requirements that actually apply to your facility.

Build the program the code assumes you already have

NFPA 110 tells you what condition the fuel system has to maintain. It does not prescribe every step of the maintenance program you should use to get there.

For a standby tank with very slow fuel turnover, we normally look at that program in three parts: testing, chemical treatment and mechanical service.

Testing tells you what is happening. At minimum, perform the fuel-quality testing required by the NFPA 110 edition and AHJ governing your facility. For tanks with long residence times, we recommend going further: establish a laboratory baseline, monitor microbial activity periodically, and sample the bottom of the tank for water and contamination. The objective is not simply to accumulate test reports. It is to see a developing problem early enough to do something about it.

Chemical treatment helps preserve the condition of fuel that has to remain in storage. A stabilizer such as Dee-Zol Life can be used to slow oxidative degradation. Bellicide is an EPA-registered antimicrobial product used to control microbial contamination. Other treatments can address specific storage problems such as dispersed moisture, sludge and corrosion. Product selection and treatment rate should be based on the condition of the fuel and the manufacturer's current directions rather than treating every tank according to the same recipe.

Mechanical service removes what chemistry cannot. This distinction becomes especially important after microbial contamination. Killing microorganisms does not magically remove the resulting biomass from the tank. If enough contamination is present, that material can migrate to filters and strainers when the generator starts.

The same is true for accumulated free water and heavy bottom sediment. Once those materials are present in significant quantities, physical removal or fuel polishing may be necessary. And if laboratory testing shows that the fuel itself has deteriorated beyond a condition that can reasonably be restored, replacement becomes the appropriate option.

The point is not that every tank needs every service every year. It is that the testing should tell you what the tank needs, and the maintenance program should be capable of responding when the condition changes.

Frequently asked questions about NFPA 110 fuel requirements

Does NFPA 110 require fuel polishing?

Not in its enforceable requirements. Section 7.9.1.3 requires that stale or contaminated fuel be remediated or replaced, 7.9.1.2 requires a supply of clean fuel to the prime mover, and the annex material discusses fuel maintenance and filtration as means of getting there. Polishing is one of the few practical methods of meeting those requirements on a tank that never turns over.

How often does NFPA 110 require generator fuel testing?

At least annually. The fuel quality test section (8.3.7 in current editions, 8.3.8 in the 2010 edition CMS enforces) requires a fuel quality test at least annually using appropriate ASTM standards. That is a floor, not a program. Most Level 1 EPSS operators add quarterly ATP microbial testing and quarterly bottom sampling, because microbial growth can go from undetectable to filter-plugging between annual tests.

Is ASTM D975 enough to satisfy the annual fuel quality test?

I would not treat that as the right question. ASTM D975 is the specification for diesel fuel, while NFPA 110 requires fuel-quality testing using appropriate ASTM standards or other guidance permitted by the applicable edition.

D975 identifies important fuel properties and the ASTM methods used to measure them, but meeting selected D975 properties does not tell you everything you need to know about a diesel inventory that has been sitting in a standby tank for several years.

Stored-fuel programs should also pay attention to problems created by storage itself, particularly water accumulation and microbial contamination. ASTM D6304 Karl Fischer testing provides much greater sensitivity to water than the conventional water-and-sediment test, while microbial monitoring can identify biological activity that ordinary fuel-property testing will not.

The appropriate test slate should therefore be based on the fuel, the engine manufacturer's requirements, the storage conditions and the NFPA 110 edition and AHJ governing the facility.

What is the 133% rule in NFPA 110?

The 133% rule comes from NFPA 110 requirements that have historically required the main fuel tank to have a minimum capacity equal to 133% of the required fuel quantity or low-fuel-sensor quantity.

That is particularly relevant to hospitals operating under the CMS code-adoption chain, because NFPA 99-2012 references NFPA 110-2010.

But don't assume that the same wording applies to every edition of NFPA 110. The fuel-capacity provisions have been revised in newer editions, including changes affecting longer-duration classifications.

The practical lesson is the same one that applies throughout NFPA 110: identify the edition adopted for your facility before using a section number or tank-sizing percentage. For facilities governed by an older edition containing the 133% requirement, the additional stored volume also contributes to the fuel-turnover problem—the facility is required to keep a substantial reserve while normal generator operation consumes very little of it.

Do I have to keep records of the annual fuel test?

Yes. NFPA 110 requires a permanent record of EPSS inspections, tests, exercising, operation, and repairs. In the 2010 edition that sits at 8.3.4, requiring records be maintained and readily available; the 2016 and later editions consolidate recordkeeping into Section 8.5 and add that records be made available to the authority having jurisdiction on request. For hospitals surveyed under Joint Commission EC.02.05.07, the fuel quality test record is part of the emergency power documentation set.

Can chemical treatment alone keep a standby tank compliant?

Not by itself. Stabilizers and biocides hold fuel condition and control microbial growth, and they do nothing about free water at the tank bottom, accumulated sludge, or dead biomass after a shock treatment. Those require mechanical removal. Fuel that is out of spec on distillation cannot be corrected chemically and has to be replaced.

How much fuel does the monthly load test actually consume?

Usually not very much compared with the amount of fuel being stored.

NFPA 110 has historically required generator sets in service to be exercised at least monthly for a minimum of 30 minutes, with diesel units meeting specified loading or exhaust-temperature conditions. But a generator's fuel consumption does not increase in a perfectly linear relationship with electrical load, so you cannot simply take 30% load and assume the engine burns exactly 30% of its full-load fuel rate.

The right way to calculate actual turnover is to use the manufacturer's fuel-consumption curve for the generator and compare the gallons consumed during routine exercising and testing with the usable gallons in the storage system.

For many large standby systems, that calculation exposes the real problem: the generator may be exercised regularly while only a small percentage of the stored fuel is consumed each year. The engine is getting exercised. Most of the fuel inventory is still sitting there.

Start with the tank you would be least comfortable explaining to a surveyor

If the annual fuel test is the beginning and end of your stored-fuel program, you may be satisfying the testing requirement without addressing the larger problem NFPA 110 is trying to prevent.

The gap between last year's sample and today's fuel condition is where water accumulation, microbial growth and fuel degradation can develop. A generator can pass its monthly exercises right up until contamination reaches the pickup, plugs a filter or interferes with fuel delivery under sustained load.

Start with the tank you know the least about.

Pull a representative sample, including an appropriate bottom sample, and establish a laboratory baseline. If the results are clean, you now have something useful to compare future results against. If they are not, you have found the problem during scheduled maintenance instead of during an outage.

Bell Performance has been working with fuel chemistry since 1909, and Bell Fuel & Tank Services handles the testing, chemical treatment and mechanical sides of standby-fuel management for hospitals, data centers and municipal fleets. Fuel Secure combines scheduled fuel testing and monitoring with the documentation needed to follow the condition of each tank over time.

The objective is simple: when the generator needs fuel, the condition of that fuel should not be a surprise.