By Rick Andrew

Imagine this scenario: You purchase a new point-of-entry (POE) water filter from your local water treatment dealer or home improvement center. You have the filter installed in your basement by a qualified individual from the dealership or a plumber. The installation is in accordance with state and local laws and regulations. A week later, you head into the basement only to discover massive water damage. The cause of the damage? A cracked filter bowl caused simply because the bowl couldn’t handle line pressure and hammering from opening and closing faucets.

The thought of this type of incident makes manufacturers, dealers and retailers cringe. Who’s liable for this water damage? In fact, the No. 1 homeowner insurance claim is for water damage with the most frequent source being the washing machine. Fortunately, this kind of scenario is much less likely to occur when certified water treatment components and devices are used. It’s no fluke, because the NSF/ANSI Drinking Water Treatment Unit (DWTU) standards used to evaluate such equipment* have specific structural integrity requirements that are met through severe tests of various types of stresses.

Three basic tests
There are three different tests that establish conformance to structural integrity requirements: a cyclic pressure test, a hydrostatic pressure test, and a burst pressure test. The specific details regarding which tests apply, the number of cycles required, and the test pressures vary according to product category, product design, manufacturers’ pressure rating, and the standard being applied (see Table 1). Requirements for other DWTU standards vary slightly from those of Standards 42 and 53, although the same three basic tests are involved in all standards.

Complete systems are categorized differently and have different certification requirements from DWTU components. For pressure-bearing components to be certified to the DWTU standards, they must not only conform to the requirements for structural integrity, but also to the requirements for material safety. Plus, for non-metallic components, structural integrity compliance includes a burst pressure test, which isn’t required for complete systems or metallic components.

In addition, the number of cycles required for the cyclic test is either 10,000 or 100,000. And the pressures for this test can be either 0-to-150 pounds per square inch gauge (psig) or 0-to-50 psig. The hydrostatic pressure test is conducted to either 1.5 times or three times the manufacturer’s maximum working pressure, depending upon the category of product. Because of these potential variables within the same base test, it’s very important to precisely categorize products prior to establishing test requirements. If there’s any question on how to classify your product, confer with your certification organization on where and what areas your device will be used in the industry. Otherwise, it will be unable to assist in a certification plan.

Test descriptions
The cyclic test is designed to simulate the opening and closing of faucets and valves throughout the lifetime of the product by cycling pressure between 0 and 150 psig. The rise in pressure must be at least one second to avoid undue stress on the test unit. A test stand with electronically controlled solenoid valves is used. This stand also incorporates an automatic cycle counter (see Figure 1).

The hydrostatic test simulates a short-term increase in static line pressure. The pressure rise must be accomplished within five minutes, but the ramp-up must not exceed 100 psig per second, again to control stress on the product. The test pressure must be maintained for 15 minutes.

The burst test establishes the required strength of the product against an instantaneous maximum pressure. The pressure rise must be accomplished within 70 seconds with the ramp-up not exceeding 100 psig per second. The test pressure must be maintained only for an instant, and then pressure is released.

Successfully passing these tests means the test product is still intact at the end of the test with no leaking detected. This differs from some other types of integrity testing that establish the conditions when failure finally occurs. Also, it’s acceptable to perform each test on a separate test unit. It’s not required the same test unit survives the full gauntlet of structural integrity tests from beginning to end. These tests require the test unit be purged of air prior to commencing pressure increase. This is necessary because air is much more compressible than water, and trapped air can cause explosions from test failures to be much more severe than they would be without air present.

Two case studies
Selecting two examples of different product types to examine and determine the application of structural integrity test requirements, let’s first examine how a faucet-mounted water filter with a maximum working pressure rating of 100 psig is tested. These are the small devices that literally hang from the end of the kitchen faucet, suspended over the sink. Because they’re installed downstream from the faucet valve, and aren’t subject to line pressure in the “off”mode, they’re considered complete systems designed for open discharge.

Structural integrity test requirements for complete systems designed for open discharge are described in the third row from the top of Table 1. As a first step, measures are taken to seal the treated water outlet of the filter element of a new test unit prior to beginning the hydrostatic pressure test. This is done per Note 2 in Table 1. The concept behind this note is that an open discharge system would only be under hydrostatic stress if the filter were plugged. And even then, those components of the system downstream from the filter media wouldn’t be subject to this stress. Given this logic, those components aren’t subject to the hydrostatic pressure test requirements and are sealed off. The system inlet is connected to the test stand, the system purged of air, and the pressure is ramped up to 150 psig for 15 minutes. The system is observed for leaks throughout the 15-minute period.

A 10,000 cyclic pressure test of 0-to-50 psig is also required. A filter system with plugged product water outlet is used for this test, just as with the hydrostatic pressure test. This test is initiated by attaching the inlet of the system to the test stand—with the system in its normal use configuration—and purging it of air. The cycling is then begun, with periodic observations for leaks throughout the course of 10,000 cycles.

As a second example, let’s look at a filter housing with a clear plastic bowl designed to accept 10-inch long × 2.5-inch diameter cartridges, that has a maximum working pressure rating of 125 psig. Because this product is sold without a cartridge, it’s a component. It fits the category of non-metallic pressure vessels having a diameter of <203 millimeters (mm) or 8 inches.

The first test to be performed is the hydrostatic pressure test. The housing inlet is connected to the test stand. The housing is purged of any air. The housing outlet is capped off with a fitting to simulate downstream closed valves. The pressure is ramped to 375 psig (which is three times the maximum working pressure) and held for 15 minutes. The system is observed for leaks throughout the 15-minute period.

A new test unit is selected for cyclic pressure testing. Again, the inlet is connected to the test stand, purged of air, and the outlet is capped with a fitting. One hundred thousand cycles of 0-to-150 psig are conducted with periodic observations for leakage made. This test will take several days to complete.

The final required test is the burst pressure test. Once again, a new test unit is connected to the test stand at the inlet, purged of air, and the outlet is capped with a fitting. The pressure is ramped to 500 psig (which is four times the maximum working pressure), and held for an instant. Note that an instant is less than a second. After an instantaneous observation of the integrity of the housing, the pressure is immediately released.

You may be thinking to yourself that these test conditions are quite harsh. After all, how much stress will that faucet-mounted filter really see? The only time it sees any pressure at all is when someone turns the faucet on, and even then it’s open discharge. And how about that housing? One hundred thousand cycles of 0-to-150 psig is a severe test of resistance to cracking. And surely the housing would never see a pressure of 500 psig in the real world.

The truth is, these test conditions are quite harsh. That’s the whole idea. And because certified manufacturer’s design and mold products conform to these harsh requirements of structural integrity, they can rest assured that these products will survive a lifetime of real world service with a significantly less chance of product failure in the field. After all, no one would appreciate a water treatment product made of very safe material, does a great job treating water and looks great, but develops a leak that leads to severe water damage after a few months of service.

About the author
Rick Andrew has been with NSF International for over five years, working with certification of residential drinking water products. He has been the technical manager of the Drinking Water Treatment Units program for over two years. His previous experience was in the area of analytical and environmental chemistry consulting. Andrew has a bachelor’s degree in chemistry and a master’s degree in business administration from the University of Michigan. He can be reached at (800) 673-6275 or email:


Comments are closed.