The Anatomy of a Refrigeration Failure: What Happens at 2:00 a.m. When Mission Critical Equipment Fails

Article contributed by Chris Wilkes, chief commercial officer at KLATU Networks, Inc.

Ultralow temperature (ULT) freezers often contain high-value biopharmaceuticals and/or irreplaceable research. When you leave at the end of the day or for the weekend, you need confidence that those assets are protected. However, based on the hours in an average work week, 76 percent of equipment failures are statistically likely to occur outside the typical 9 to 5 schedule. This explains why it is so common for facility managers or researchers to be alerted in the middle of the night or over a long weekend of a failure.

Anatomy of a refrigeration failure

Unlike many pieces of laboratory equipment, cold storage is always in use. Maintaining temperature, especially for ULT freezers (-80°C / -112°F), is hard on its componentry and even the most reliable units will fail, typically 10 to 15 percent of a fleet each year. So, on a Friday afternoon when a partially loaded ULT freezer’s compressor stops working, the temperature gradually rises until it reaches the alarm setpoint of -60oC which can be 12 hours later. If there is a monitoring system, the alerts only sound after the temperatures have reached the alarm thresholds. Often hours after the system has failed. Because of this, a small window of time remains to move and secure the samples to standby storage before product loss occurs.

Protecting assets with predictive analytics

While waiting for a failure to occur and then fixing it has been the industry status quo, predictive monitoring – or Monitoring 2.0 – has been able to help equipment managers detect patterns and send alerts hours, days or weeks before a freezer fails. Monitoring 2.0 systems utilize statistical models and advanced algorithms to determine the health of the units.

Predict-and-prevent in action

To understand how predictive monitoring works its useful to look at the common causes of refrigeration failures and how Monitoring 2.0 behaves differently from traditional laboratory monitoring systems.

Advertisement

Compressor or electrical component failures: Units with ice covered gaskets, clogged condenser filters ora slow refrigerant leak will ultimately lead to early failure of the compressor or electrical components. While these stressed units maintain their setpoint temperature, the refrigeration systemis forced to run continuously which shortens its useful life, typically by 20 percent.

Traditional monitoring only issues an alarm once the unithas failed and its contents warm beyond the setpoint, often resulting in product loss or degradation of quality or efficacy. However, predictive monitoring can analyze historical data and identify when the unit becomes stressed. Flagging this early, can help the facility management team react before an issue has occurred.

Equipment misuse or overuse: When there is a high amount of door openings or excessive heat loading, freezers will experience early failures because of high on-time. Chronic high on-time in the context of HVAC related issues, cabinet temperature, power consumption and door openings are flagged by predictive algorithms as abnormal—providing alerts days or weeks before a potential failure. Leveraging this predictive technology allows users to get ahead of maintenance issues, further extending the useful life of the units.

Ineffective repairs: Independent research from KLATU Networks estimates that about one-third of all repairs are not effective 30 days after the repair event or six-months later. Monitoring 2.0 systems can compare performance data pre- and post-repair to determine how much pre-failure performance is restored after the repair. In addition, the system can ensure the repair is persistent.

Because your assets are often irreplaceable, waiting for an alarm at 2 a.m. is often too little, too late. Transitioning to predictive monitoring can help you move from a fail-and-fix to predictive maintenance (PdM) practices. Not only can Monitoring 2.0 flag assets immediately when they show signs of mechanical stress helping prevent failures, but your team can also schedule maintenance based on changes in health-status and condition minimizing the 2 a.m. crisis. Making the shift to PdM is not only beneficial for your precious research and biopharmaceuticals, but it can also help your bottom line by reducing after-hours emergency responses, lowering average repair costs, improving energy efficiency and ultimately extending your asset service life.