Key Considerations in Fire Pump Design and Installation

Fire pumps are a critical component of a building's sprinkler system in settings where the water supply is insufficient to provide the pressure needed to keep the water flowing to all the sprinkler heads. Fire pump intakes are connected to underground public water supply piping, or a tank or reservoir located onsite, to provide water flow at a higher pressure to the sprinkler system risers and hose standpipes. 

SGB Product Page

Fire pumps are activated when the water pressure in the system drops below a certain threshold. For example, this can happen when one or more of the sprinkler heads are exposed to heat exceeding the temperature they were designed for or when other firefighting connections are opened, causing pressure in the mainline to fall.

While trained fire protection professionals must perform the design and installation of fire pumps, this post will provide an overview of the different types of fire pumps in use today and summarize the factors considered when choosing a pump. 

Does Your Sprinkler System Require a Fire Pump?

Not all sprinkler systems require a fire pump. However, those installed in most commercial buildings do because the pressures needed to ensure adequate flow to all parts of the system usually exceeds what a standard public water supply can provide. When the sprinkler system is connected to a static water source, a fire pump is always required. 

If your building doesn't have a fire pump and annual testing of your existing sprinkler system reveals a pressure issue, you may be required to install one, depending on the nature of the problem.         

Because fire pumps are an integral part of so many sprinkler systems, there are many sections of the National Fire Protection Association (NFPA) 13 Standard for the Installation of Sprinkler Systems that reference them (e.g., pump room requirements, testing, etc.).  

However, the requirements pertaining specifically to the actual design and installation of fire pumps are found in the National Fire Protection Association (NFPA) 20 Standard for the Installation of Stationary Pumps for Fire Protection. In addition, if your system is supplied with an onsite tank, you will find additional requirements for their installation and maintenance in NFPA 22 Standard for Water Tanks for Private Fire Protection.

Different Types of Fire Pumps and their Applications

Positive Displacement Pumps

Positive displacement pumps create flow pressure by repeatedly capturing a fixed volume of water into the pump's internal chamber using suction to compress it before releasing it out of the discharge valve. These pumps are not as common because their flow volume is more limited than that which centrifugal pumps can provide and are typically only used with water mist or foam-water systems.

Centrifugal Pumps

Centrifugal fire pumps are the most common type of pump used in commercial buildings today. This type of pump creates pressure through centrifugal force with an internal impeller. When the pump is activated, water flows into the center of the impeller, which rotates at high speed to drive the water to the discharge valve creating pressure by pushing high volumes of water through the valve. There are several different types of centrifugal pumps, which are described below in terms of some considerations that might factor into the selection of a fire pump for your building.

*gpm = gallons per minute

Choosing the Right Type of Pump Driver 

The most common types of fire pumps in use today are electric motor-driven and diesel engine-driven fire pumps.

Electric-driven pumps are always preferable where possible because electric motors are more compact and require fewer mechanical parts, making them very cost-effective. They also have fewer negative environmental impacts. Electric-driven pumps can run off a utility connection, a generator, or other power source approved by your AHJ. 

Diesel-driven pumps are typically used where the power grid is unreliable or unable to handle the load of an electric-driven pump. They are also used where there is a lack of emergency power available, such as a generator.   

You can expect more regulations for diesel-driven pumps than electric pumps due to the extra steps necessary to handle the fuel system, cooling, exhaust, vibration, etc. For example, diesel-driven pumps require additional space for fuel tank storage and must be located in a separate enclosure or room with direct access to the exterior.

How Fire Pumps are Sized

The size of the fire pump needed is usually determined during the design of a sprinkler system. When designing a sprinkler system, a hazard analysis must first be conducted to determine theoccupancy classifications, which are defined based on the expected fire hazards in different spaces within the building. This analysis determines the number of sprinkler heads required. And, while the number of sprinkler heads served is an important factor to consider in sizing a fire pump, the size required is based on the most hydraulically demanding area of the entire sprinkler system, which may or may not be the area with the most sprinkler heads.  

Determining the most hydraulically demanding area in the system is a highly complex calculation that requires qualified fire protection professionals to conduct. Factors that go into calculating it include:

• Distance — The building's footprint will determine how far water must travel laterally to reach the farthest sprinklers.
• Elevation — The designer must take into account how many floors the building has. Due to the effects of gravity, more pressure will be required to push water to sprinklers on higher floors. 
• Occupancy Classification — It is important to know if there are hazards within the building that might require more water to suppress a fire and where they are. This is why the hazard analysis must be conducted before sizing the fire pump. 

Correctly sizing the fire pump is made even more difficult because units expressed in GPM cannot be directly converted into PSI. Thus, while the capacity of the sprinkler system is defined in terms of gallons per minute (GPM) -- the amount of water flowing through the system, fire pumps are typically sized by pressure range, pounds per square inch (PSI).

Questions About Fire Pumps?

Welcome to the second and what (I hope), will be the final part of my series on this humble but critical component the:

Centrifugal Pump Wear Ring

Those of us who deal with the specification, design, manufacture and troubleshooting of these machines on a daily basis often forget that this simple part belies a lot of complexity as wear rings have major impacts on the functioning of the pump. In this Part 2 we'll cover the secondary purposes of the wear ring as well as the important topic of material selection and wear ring fitment.

Along the way I've tried to cover all of the questions raised by the many comments from Part 1 of the series. My apologies in advance if I missed any, I'm certain you'll let me know in the comments below.

***I strongly recommend that if you arrived directly at this article that you take the time to read Part 1, otherwise the topics below might seem disjointed***

Secondary purposes for Wear Rings

1. The Lomakin Effect

On multistage pumps the radial clearance wear rings provide support for the rotor via the Lomakin effect. I'm not going to explain this effect in detail here (in the interests of not making this a 10 Part Series), but you can read about it at these links

• An End-User’s Guide to Centrifugal Pump Rotordynamics
• Pump Rotordynamics Made Simple

Essentially each wear ring acts like a fluid film journal bearing - providing support stiffness and damping for the rotor. Without this effect it would not be possible for multistage pumps to operate without damaging rotor movement, continuous rubbing and rapid wear. The crucial thing to note is that the strength of the Lomakin effect is a function of:

1. The pressure drop across the wear ring - which is a function of fluid density
2. The clearance of the wear ring
3. Other wear ring geometry - features like diameter, length, grooves etc.

Since 1 and 3 are generally fixed when a size of pump is selected for a service, varying the wear ring clearance is usually the design variable we can most easily utilize to modify the pump rotor behavior.

In computing the wear ring stiffness and damping resulting from the Lomakin effect, the key things to understand are that

• Increasing the wear ring clearance is detrimental (usually highly detrimental), to the resulting stiffness and damping.
• Pump OEMs need to demonstrate acceptable lateral rotordynamic behavior not only when the pump is new but also up to the point it is taken out of service for routine repair. This means whatever "New" wear ring clearance is utilized, the rotordynamic behavior has to be acceptable at 2x "New" clearances which signify the pump in a "Worn" condition.

So it follows if your new pump starts off with a large wear ring clearance (such as API 610), getting acceptable lateral rotordynamic behavior at 2x those clearances can be challenging.

Big clearances + Thin fluid = Rotordynamic fail

For fluids with densities close to water, it generally isn't a problem to accommodate API 610 and larger wear ring clearances. The pump OEM can usually achieve acceptable lateral rotordynamics. However as the density of the pumped fluid is reduced, very quickly it becomes impossible.

To demonstrate this I pulled some data from a recent 3 stage BB5 lateral rotordynamic analysis we performed. The customer in this case had insisted that the pump should have API 610 clearances. The table below shows the direct stiffness and damping that resulted.

In the above example you can see the huge variation in resulting stiffness and damping as a result of the clearance change. As the pump wears the available stiffness drops by 2/3rds...

For this reason most pump OEMs will offer multistage pumps with clearances less than API 610. For multistage pumps handling very low density fluids (below about 0.6 Specific Gravity), the use of further reduced clearances (and non-metallic wear rings) is a routine practice.

If you are looking for more details, kindly visit Horizontal Single-Case Pump.

This brings us to a customer error that I've seen multiple times over the years:

Some customer specifications will require the use of API 610 clearances regardless of the pump type or fluid being pumped. Explaining to them why their specification violates the laws of physics tends to be a painful discussion. Please don't be one of those customers...

2. Axial thrust management

Wear rings (and wear surfaces) play an important role in the management of the axial thrust developed inside the pump. This is due to their role of separating high and low pressure regions within the pump. On single and two stage pumps the pump designer will typically vary the diameter and placement of the wear rings in order to achieve acceptable balancing of the axial forces acting on the impeller.

This can be visualized on the diagram below. By vary the wear ring diameters of the eye and hub side wear rings, the area over which the green and red pressures act and hence the resultant axial thrust can be controlled.

I've previously done a complete series on axial thrust in centrifugal pumps so the reader is recommended to refer to at least Part 1 of the series for the basics of how axial thrust develops in the pump and how it can be managed.

• Axial Thrust Part 1 (The basics are covered here)
• Axial Thrust Part 2
• Axial Thrust Part 3

Material Selection

There are several things which guide or control for wear ring material selection:

Specifications

API 610 and some customers will have specifications defining acceptable wear ring materials based on their experience in a specific service. For example API 610 Table H1 identifies the wear ring materials to be used for a specific material class. (Note that API also allows certain non-metallic wear rings in accordance with Table H3).

API 610 also requires that "hardenable" wear ring materials have hardness difference of 50 Brinell between the rotating and stationary surfaces (refer to clause 6.7.2).

For applications with Hydrogen Sulfide (H2S), NACE MR or MR will likely be invoked. In these cases API 610 will limit the hardness of the rotating wear ring to less than Rockwell C (HRC) 22. Clearly this has implications when trying to achieve a 50 Brinell differential hardness and the pump OEM will have a specific material combination to meet these requirements (clause 6.12.1.14.1)

Tribology

While the ideal state is that wear rings operate at all times with separation between the stationary and rotating surfaces, the reality is that contact between wear rings will happen...

• Transient events such as thermal shock, water hammer etc. can result in momentary contact
• Rotor misalignment (from coupling loads, poor driver alignment, external piping loads etc.) can result in varying degrees of contact
• In most multistage pumps the rotor wear rings will contact the stationary rings during startup and shutdown

Consequently it is important to pick wear ring materials that can tolerate at least some degree of contact. There are literally thousands of possible material combinations and this article is not intended to replace expert advice from your pump OEM. However I've listed below a few pairings and their overall scoring on a few key criteria: galling resistance, resistance to wear by solids and applicability / cost.

Please use the above table as a starting point for a discussion and not as an immutable reference or shibboleth.

Corrosion

If a wear ring corrodes, clearly that will have a negative impact on its performance since the clearance will increase. The following are considerations when selecting a wear ring material:

1. The wear ring material should preferably be more corrosion resistant (noble) than the pump casing. Otherwise corrosion will tend to be concentrated on the small wear ring component
2. Be aware of Flow Accelerated Corrosion (FAC). Specifically when a certain velocity threshold is reached, the corrosion rate rapidly increases. Because the velocity threshold is dependent both on material and fluid there are no simple rules and you should always consult a materials expert. The graphic below shows the expected corrosion rates for different velocities of seawater for a few materials to give you an idea of how much this can vary.
3. In warm seawater applications where the fluid can become stagnant and depleted of oxygen, pitting corrosion of typical materials such as austenitic, duplex and (when polluted water is present) even super duplex stainless steels is a significant risk. In these cases the resistance of the material to pitting needs to be checked for the expected conditions - especially anywhere fluid is present in the pump but not moving.

Non-metallic wear rings have their own specific considerations

The primary concern for me is that they wear out quickly in dirty or contaminated fluids (see also the table above).

Simon's Rule Of Thumb is that I would only allow the use of non-metallic wear rings when the suspended solids content of the fluid is < 300ppm.

As I write this I can already hear some supplier of non-metallics typing furiously in the comments section about how their "proprietary material" is the best in dirty services. Sure, there are exceptions to the limit stated above, however they should be reviewed in detail with the supplier and the supplier should warranty that their material will deliver acceptable life.

The secondary (design specific) concern is that the coefficient of thermal expansion of these materials is very different from any metal. Hence for applications where the temperature differs significantly from ambient, it is very important that the fit between the wear ring and the component into which they are mounted is correctly Engineered.

Wear Ring Fitment

A few issues merit consideration here. Many people see wear ring with set screws between the ring and the impeller or casing and assume incorrectly that it is the set screws alone that provide the retention.

In actuality it is the fit and shouldering between the wear ring and mounting component that provides the primary retention. The set screws (or adhesive or tack welding or pinning) is just a secondary assurance and to provide some anti rotation functionality.

So with that said here are the issues to consider:

1. If the pressure inside the pump during operation tends to push the wear ring off its register fit, an interference fit is needed to assure retention.
2. If the pressure inside the pump during operation tends to push the wear ring onto its register fit, a close clearance fit is generally ok.
3. When the wear ring material differs from the casing or impeller material, the effect of temperature and the resulting thermal expansion need to be computed. Thermal transients can further exacerbate this as the wear ring tends to react to the thermal transient before the larger mass of the casing or impeller. I've seen more than a few cases of wear rings becoming loose and displaced due to this.
4. For pump impellers operating at high speed, the centrifugal stress and deflection of the wear ring needs to be computed and allowed for in the design. For this reason very high speed impellers will often have integral wear surfaces rather than separate wear rings as the combination of high pressure pushing the ring off the impeller + deflection from centrifugal stress become problematic.

Now lets talk about the secondary fixing arrangements. Set screws tend to be the most common but there are others. I've made a summary table of the ones I can think of and their positives / negatives. I fully expect the active commentators to suggest a few more obscure ones ;)

Effect on NPSHR

There is not a huge amount of study and data on this. However we know in general that the following will affect it:

1. Increasing the wear ring clearance beyond a certain value will have a noticable impact on NPSHR
2. Pumps with higher suction specific speeds (high Nss), appear to be more affected by increasing wear ring clearance

In I was part of the group that performed a study looking at the effects of wear ring clearance.

• You can read my summary of the findings of the study in this article
• Alternatively you can read the whole paper here

The study found that the increase in NPSHR at "end of life clearances" ranged from 5% to 21% depending on the impeller suction specific speed (Nss)

Alternatives to Wear Rings

We talked about this in Part 1 in passing. The primary type of impeller that utilizes a Wear Surface as opposed to a wear ring is the open impeller. I've shown a generic example below where leakage of fluid from the high to low pressure zone is controlled by a close axial or conical clearance between the impeller vanes and a stationary counter surface mounted in the casing.

Another common alternative is where instead of a hub side wear ring to help manage axial thrust, the impeller is fitted with pump out vanes that form a close axial clearance with the pump casing or cover. You will often see these on pumps handling fluids with high solids content as the vanes are helpful in reducing the tendency of solids to accumulate around the mechanical seal.

The pump out vanes act to modify the pressure profile on the rear of the impeller reducing (to an extent), the resulting axial thrust.

Ok that's all for Part 2. Thank you to anyone who managed to get through to this point. If I've missed any of your previous questions or you feel I didn't cover something correctly or in sufficient detail, please let me know in the comments.

As always your comments (whether agreeing or disagreeing), are most gratefully received.

Are you interested in learning more about Horizontal-Shaft Axial Flow Pump? Contact us today to secure an expert consultation!