Reporting Release Rates: How to find the middle ground
“We can actually analyze an entire system of a facility and see in what scenarios a release could occur and estimate how much ammonia would actually be released.”
– Bent Wiencke, Ammonia RefrigerationFoundation-funded research project leader
“When covered facilities here in the U.S. have an accidental release of ammonia, they need to quickly determine if the release amount has or will exceed the reportable quantity of 100 pounds,” said Douglas Reindl, a professor at the University of Wisconsin and director of the IRC. “If the end-user does not promptly report (they have 15 minutes from when the leak was discovered to make this determination), they are subject to hefty fines.”
“Facilities want to accurately report but not over-report to avoid unwarranted regulatory scrutiny, but if you have another situation where there is an accidental release and the neighbors call 911 or the fire department triggering an external response, that will get regulatory attention whether or not the reportable quantity is reached,” Reindl said.
Bent Wiencke, who has more than 30 years of experience in the industry, led the Ammonia Refrigeration Foundation-funded research project to prepare comprehensive guidance on estimating ammonia leak rates. He said the research has been challenging due to the complexity of the methodologies, and the guidance document itself quickly exceeded 100 pages in length.
“Some of the equations used for estimating refrigerant release rates are difficult to solve and detailed explanations had to be provided to remove ambiguity. Furthermore, if we condense this document down to a few pages, regulators could challenge the validity of the document and ask a lot of questions on the method’s basis,” Wiencke said.
This will be the first formal guideline for estimating accidental release rates specific to the ammonia refrigeration industry. “There are various leak calculators and spreadsheets on leak rates one can find on the internet, but you don’t necessarily know where they came from, what they are based on, or their underlying limitations, so some regulators have had issues with facilities using these various leak calculators,” Wiencke said. “If you tell me, your total release amount is just below the threshold limit of 100 pounds but you can’t tell me how you got to that number, I will assume it is simply based on a guess and not based on any sound and established methodologies.”
The forthcoming IIAR accidental release calculation guideline will cover single liquid releases with saturated and sub-cooled liquid and gaseous releases with saturated vapor and superheated vapor. The guideline allows users to estimate accidental refrigerant release rates based on some basic inputs that include upstream refrigerant pressure, hole size, and the relevant geometric details that would include the location (where the pipe or vessel wall experiences loss of containment, where a nozzle or valve is sheared off, or a hole is drilled into the vessel or pipe wall, or a pipe section comprising of a pipe section, fittings and valve is severed). However, it isn’t designed to cover severe catastrophic releases that are transient, Wiencke said.
USEFUL RESOURCES
IIAR’s release calculation guidance will be a print document, but IIAR is planning to provide an Excel-based spreadsheet that will facilitate solving those complex equations.
“I started realizing very quickly that these equations are very cumbersome to solve, and all of the guideline’s users would have to ‘program’ the equations themselves, which is repetitive and prone to error for each user,” Wiencke said, adding that providing a single source spreadsheet will make it easier for people using the tool and allow them to more quickly calculate release rates. “You plug in hole size, pressure and if the liquid is saturated or subcooled or the vapor is saturated or superheated and within a minute, you have an estimate. If you started entirely from scratch, it would probably take several hours to solve some of the more difficult relevant equations.”
The guidance can also help with planning. “With ammonia, we do Process Hazard Analysis and these analyses often look at worst-case scenarios involving accidental releases at various locations throughout a system. The spreadsheet will be handy because we can do an analysis of our systems and quickly determine how severe could a worst-case scenario be by performing quick calculations using the Excel spreadsheet template,” Wiencke said. “We can actually analyze an entire system of a facility and see in what scenarios a release could occur and estimate how much ammonia would actually be released.”
While these formal guidelines on estimating accidental leak rates and the research IIAR is doing on machine room ventilation are standalone projects, there is a connection, Wiencke said. “They are independent, but they are covering some of the same issues we have in the industry,” he said. “The machinery room ventilation used simulations based on certain accidental release assumptions and you have to make sure the corresponding release rate calculations are correct. So, I collaborated with the primary researcher on that project to make sure we’re internally consistent on our accidental release rates.”
FUGITIVE RELEASES
Independent of the ARF project on ammonia release rate determination, the ammonia refrigeration industry has also been examining fugitive emissions of ammonia, the unintended and undetected loss of refrigerant from a refrigeration system that occurs intermittently or continuously, which can be difficult to calculate.
Eric Smith, vice president, and technical director at IIAR said it is important to distinguish between release rates resulting from fugitive emissions and an accidental uncontrolled release. “What is interesting is that all refrigeration systems seem to have some amount of fugitive emissions, whether that be from service venting or very small leaks from the system,” he said, adding that it is not uncommon for some facilities to lose between 6 and 20 percent of the total refrigerant charge from their system on an annualized basis, but fugitive emissions from ammonia refrigeration systems tend to be lower compared to other refrigerant-based systems because of ammonia’s odor.
“Refrigerated facilities have improved their follow-up based on personnel reporting ammonia odor and refrigeration staff are taking prompt action to identify the leak source and initiating repairs,” Smith said. “Refrigeration systems using fluorochemical refrigerants do not have this same advantage because these refrigerants do not have a distinct odor.”
Reindl said the Industrial Refrigeration Consortium has just finished up a two-year project focused on fugitive emissions from industrial ammonia refrigeration systems, a project funded by EPA Region 5. “One of the goals of the project was to gain a better understanding of the underlying causes for some facilities having to add refrigerant to their systems more frequently and others less frequently.”
During IIAR’s virtual conference, Marc Claas, a research engineer with the Industrial Refrigeration Consortium, discussed results from the fugitive emissions project. “For more than a decade we have informally gathered anecdotal evidence of refrigerated facilities adding between 1 percent of the total system charge annually to more than 100 percent of the total system charge. Data gathered in the field as part of this project is convincing us that the higher end of this range of refrigerant losses for ammonia systems is not attributable to fugitive emissions,” Claas noted.
As part of the fugitive emissions project, the IRC project created a dynamic inventory tracking tool in Excel to help refrigerated facilities more quickly be alerted of potential refrigerant losses due to fugitive emissions. Claas described the inputs to the tool as the high-pressure receiver (HPR) dimensions and orientation, which allows the tool to calculate the vessel’s volume. From there, users begin entering data that includes the date, current HPR liquid level, and the saturation pressure or temperature of the refrigerant in the HPR.
Using refrigerant properties included within the tool, the total charge of refrigerant residing in the high-pressure receiver is calculated. Once a number of data points over a period of time are compiled, the tool plots the HPR charge data to reveal trends in refrigerant inventory.
If the refrigerant loss rate is high, the tool flags facility staff to take action and find the source of refrigerant loss. Claas noted that “at first, the refrigerant inventory data appears a bit random but as users compile more data over a period of time, trends will emerge and the tool can estimate an annual refrigerant loss rate.”
“What is interesting is that all refrigeration systems seem to have some amount of fugitive emissions, whether that be from service venting or very small leaks from the system,”
– Eric Smith, vice president, and technical director at IIAR
“Ideally you take a charge calculation at some regular interval, daily or weekly, and trend that charge calc over time to see how much refrigerant is leaving the system,” Claas said. “Based on our experience thus far, tracking a single uncontrolled level vessel is enough to trend the overall system charge, with certain caveats considered of course.”
For facilities going several years between “topping off” their system with ammonia, this tracking tool can provide indications of accelerating refrigerant losses and prompt follow-up. The results of the fugitive emissions project show odor screening for ammonia is a good place to start in reducing fugitive emissions. It is also important to investigate potential releases into water or secondary fluids, as these will have low or no odor.
“One problem with fugitive releases is that on an extremely large system, fugitive releases could occur from many small sources on a system,” Smith said. While a leak from any single point on the system could be much less than 100 pounds per day, combining all the fugitive releases could produce a total exceeding 100 pounds a day, which is a reportable amount. “That is a situation that should be clarified with regulators because the rules are intended to report accidental releases rather than fugitive leaks,” he explained.
