There are different criteria for UV systems based upon the target use. Currently, NSF International certifies the performance of water treatment and distribution products for drinking water, industrial and waste water, and recreational water. Below are the most important NSF criteria addressing UV systems:

• NSF and EPA ETV Generic Protocol for the Development of Test/Quality Assurance Plans for Validation of UV Reactors

• NSF/ANSI 50: Equipment for Pools, Spas, Hot Tubs, and Other Recreational Water Facilities

• NSF/ANSI 55: Ultraviolet Microbiological Water Treatment Systems

• NSF/ANSI 61: Drinking Water System Components - Health Effects

The ETV protocol and NSF/ANSI Standard 50 were recently harmonized to form more efficient requirements. The ETV protocol fully complies with the requirements in the US EPA UV DGM.

It’s important to note that NSF 50 references the NSF EPA ETV protocol, which validates UV systems for log inactivation of cryptosporidium. NSF 50 contains design and performance testing requirements for corrosion resistance, material health safety, burst pressure tests, pressure loss tests, 3,000-hour life testing, three-log bacterial disinfection efficacy tests and product marking, installation and use instructions.

NSF 55 is used to evaluate UV treatment systems for low flows such as POU and residential POE products. It includes two levels of performance testing. NSF 55 is typically required for~35 gpm and lower flow POE/POU UV systems.

NSF 61 is a material health effects safety standard and does not contain functional performance or microbiological testing requirements. It is used to assess if drinking water contact items for organic, inorganic, metals, and the like are leaching contamination risks.

In 2006, the United States Environmental Protection Agency issued the Ultraviolet Disinfection Guidance Manual. The USEPAUVDGM provides criteria to help cities and water utilities with site assessment, system design, installation, testing and operation of UV systems. NSF worked with the U.S. EPA and state drinking water authorities to create a set of testing criteria based on USEPAUVDGM. From this effort, NSF published the NSF and EPA ETV Generic Protocol for the Development of Test/Quality Assurance Plans for Validation of UV Reactors in 2009 (revised in 2011). The ETV UV protocol is consistent with the USEPAUVDGM. It is “the” standard for UV performance testing.

Newly popular sprayparks, however, were designed differently from pools and spas and were not held to NSF 50’s traditional requirements for equipment. Many spraypark cryptosporidiosis outbreaks occurred related to facilities that were not using NSF-certified UV systems. That is changing. Now states that have a requirement for UV systems mandate that they be tested and certified to NSF 50, and for higher risk uses (such as spraypads) to achieve a minimum three-log (99.9 percent) inactivation of cryptosporidium. Some states have required NSF 50 certification and a delivered UV reduction equivalent dose of 40 mJ/cm2, for higher risk spray parks.

NSF 50 requires a minimum of three-log cryptosporidium inactivation performance (greater performance such as 40mJ/cm2 or four-log also is accepted) for UV system manufacturers that make product performance claims related to cyst inactivation. NSF 50 also cites the EPA ETV UV protocol and incorporated new testing requirements that included a more specific and repeatable derivative of LT2ESWTR and UVDGM, as well as requirements from DVGW W-294 and ONORM 5873 (European criteria for UV system evaluation).

Due to this harmonized approach, UV systems may be easily evaluated for drinking water and recreational market uses.

NSF certification for cryptosporidium inactivation includes these important UV reactor validation steps:

1 Obtain technical specifications for the system from the manufacturer

2 Assess the UV sensors

3 Perform collimated beam laboratory bench scale testing

4 Perform full-scale reactor testing

5 Calculate the reduction equivalent dose (RED).

6 Adjust the RED for uncertainty in UV dose.

7 Calculate a validated dose for cryptosporidium to show a minimum three-log reduction

The ETV protocol mandated the use of germicidal-only sensors, allowed DVGW- or OONORM-certified sensors, rejected duty sensors not within 5 percent of reference sensors, mandated radiometer agreement within 5 percent, required pipe configuration with a 90-degree bend for testing, specified conditions for performing organism stability testing, and specified the minimum number of sensors for lamps.

Other improvements resulting from ETV stakeholder reviews include precise timing of sample collection, a procedure for allowing test organisms other than bacteriophage MS2, frequency of flowmeter calibration, and frequency of reactor blanks.

Certification is not based on a single validation, but requires ongoing conformity assessment to ensure compliance with USEPAUVDGM and manufacturer performance claims. Unlike a validation (which is a one-time test report), certification also addresses production changes over time through annual onsite audits of the entire system. Section 5.13 of USEPAUVDGM specifies which components and parameters require evaluation to determine the need for revalidation testing if they are modified. Only NSF monitors these issues, and that’s why NSF certification is so important and much more valuable than a single validation. Certification includes that validation test along with ongoing conformity assessment.

Following are some examples of UV changes that might affect performance and hence require revalidation testing:

• Lamp type, gas mixture, power density, arc length, end reflectors, and UV transmittance of the lamp envelope between 185 – 400 nm

• Sleeve/shield thickness, material type and spectral absorbance

• Sensor type, supplier, spectral response and location within reactor relative to lamps

• Ballast operating voltage, current, frequency and waveform

• Wetted geometry, reactor dimensions, design and hydraulics

Certification allows a company to control which proprietary information it releases by limiting public display of proprietary information while disclosing nonconfidential details. The full report developed under NSF certification is NOT placed in the public domain.

When NSF certifies UV systems, it enables worst-case unit testing. Per NSF 50, NSF selects two (or more) worst-case designs for biodosimetry testing. This testing is conducted in accordance with NSF/ANSI Standard 50 and the NSF/EPA ETV UV protocol (i.e., the U.S. EPA UV DGM compliance validation testing). With NSF, the manufacturer may select from three levels of testing detail to address data needs and use:

1 set-point,

2 set-line, or

3 calculated dose types of validation testing

Set-point testing involves the minimum set of testing required to comply with NSF/ANSI 50 and the NSF/EPA ETV UV protocol. With this testing, the validation report is concise but addresses all the critical needs of both public health authorities and users and is perfectly suited for manufacturers whose system users will typically run the system at a certain target flow rate and intensity. This “set & forget” covers most recreational water facility needs.

A more data-intensive approach is set-line testing. It is used to analyze a UV system running at one flow rate, but includes testing at a few different power levels and different UVTs. Set-line testing essentially triples the number of flow rates tested from set-point testing and gives more verified conditions.

Lastly, calculated dose testing goes even further with the number of conditions that are modified. The resulting larger data set can have greater cost implications for a manufacturer, but certain applications merit this testing. The manufacturer obtains a report that shows different conditions under which the system may be operated and still deliver a certain “dose.” This can be useful for operators of larger UV systems that consume a lot of energy. In certain drinking water treatment operations, a manager may be able to reduce power or increase flow and still operate within acceptable conditions.

Regardless of which type of testing is chosen, a detailed report is created for the manufacturer. This can be used to meet the needs of a regulatory authority, facility operator, drinking water utility operator or the manufacturer.