FrigoBoss Case Study: Incipient Fault Detection of Refrigerant Leakage and other problems

Introduction

One of our certified partners and installers brought to our attention the following problems a client of his was facing. The client is a distributor of frozen and cold goods to local supermarkets and food stores, located in central Greece. He owns a facility with 10 cold rooms. Each cold room has an approximate volume of 600m3. The facility was old and was running on the original machinery (compressors, motors, etc.) as well as on their obsolete control systems. For the past few years, the client tried to operate his facility as much unattended as possible. The combination of various problems with the existing infrastructure itself, the frequent failures of equipment and system components without any prior warning and the many frequent power outages in the area has led to significant losses of stored goods, and thus great loss of profit and financial penalties as a result of the damaged goods. Furthermore, the client’s suppliers and customers required the proof of traceability and temperature monitoring of the stored goods to ensure their quality and to avoid health hazards.

The requirements

The client asked for the feasibility of a solution that could first of all provide the necessary regulatory compliance for temperature recording and monitoring, as well as, to have a degree of early warning for failures and real-time monitoring and alarming, and all this at a minimum available budget.

Achieving regulatory compliance and operational excellence with FrigoBoss in one fell swoop

The core function of the FrigoBoss system is to provide full regulatory compliance to EN 12830:2018 and HACCP requirements. This is achieved by temperature recording and monitoring functions of the FrigoBoss system, utilizing only the dedicated temperature transmitters of the type AP-150-901 coming with the system. The available budget of the customer was very limited and did not allow for the installation of additional sensors and instrumentation, like power meters, pressure transmitters, etc., that would lead to a total performance monitoring solution for its installation. Thus, we had the challenge to design a solution that would meet the temperature recording regulatory requirements and at the same time to provide the early warning of the failures, the fault prediction and the fault isolation capabilities utilizing only the temperature transmitters. So, this task became also a number of sensors optimization challenge.

We first started by optimizing the number of sensors required for each cold room by specifying the optimal installation locations of the temperature transmitters within each cold room, in order to accurately capture the temperature measurements required to meet the dual requirement described above. The whole installation was covered with a total of 54 strategically installed temperature transmitters around the facility. The highly competitive list price of our AP-150-901 temperature transmitters enabled us to meet and exceed the customer’s requirements and to be able to install a total of 54 temperature transmitters. The flexible and modular architecture of the FrigoBoss system, along with the preparation from our side of detailed installation drawings and installation instructions allowed for very fast installation time and trouble-free commissioning of the whole system.

Upon the FrigoBoss system’s commissioning and start-up, the goal for regulatory compliance was achieved. In order to meet fault finding, fault isolation and early warning of failures requirements, one of the most powerful and advanced features of FrigoBoss, one of its aces upon its sleeve, was called into play. This feature is the embedded powerful algorithms for automatic incipient fault detection, fault prediction and fault isolation. This approach relies on a combination of physics-based and data-driven models. This approach is something completely different than simple, classical, rules-based alarming. The effectiveness of this powerful feature will be demonstrated in the following analysis. A snapshot of the measurement trends of the installation will be used to showcase 3 different faults successfully identified using the FrigoBoss system.

Fault Cases Identified by FrigoBoss

Fault Case No1: Refrigerant leakage fault detection and isolation utilizing only temperature measurements

About 20 days after the installation of the FrigoBoss system at the client’s premises, a serious failure occurred at the refrigeration system that would have otherwise gone undetected until it was too late, due to the nature of this fault and the way the facility was operated. This fault was refrigerant leakage from the refrigeration circuit.

The technical challenge was to identify the fault right at its initiation, as well as to accurately distinguish it in relation to other faults presenting the same symptoms with respect to temperature measurements, as only temperature transmitters were utilized in this implementation of FrigoBoss. Figure 1 is a snapshot of the installation’s historical temperature trends showing the steps of how FrigoBoss successfully picked up this fault.

Figure 1: Refrigerant Leakage detection and system behaviour from the scope of temperature measurements.

Point 1: Time of the initiation of the refrigerant leakage from the refrigeration piping circuit.

Point 2: The combined, physics-based, and data-driven modelling algorithm of FrigoBoss picked up the fault and distinguished it successfully between faults presenting the same symptoms with respect to temperature measurements (like compressor motor failure, electrical power supply failure, etc.). Critical role to the successful fault isolation played the strategically placed temperature across the individual cold rooms as well as on the compressor motor casings. Figure 1 shows the temperature measurements of Room 1 and Room 2 of the temperature transmitters installed on the respective evaporators of these two cold rooms, as well as of the two compressor motors. The sensor installation locations were identified during the design stage of this project. At the same time, the system generated the respective alarms and automatically notified via SMS and email the designated users, in this case, the owner and the technical support external partner.

Point 3: The problem was resolved by the timely action of the customer’s technical support team, due to the early warning and detection of the failure.

The prompt actions taken due to the incipient fault detection and early warning provided by FrigoBoss enabled the customer to save the goods stored in the utilized cold rooms of his facility, preventing a total loss of goods valued 100000 EUR, along with any penalties applying. With this single action, FrigoBoss has literally achieved an instantaneous ROI many times over. In the past, this kind of failure in this facility would have gone completely unnoticed until it was too late, and this was one of the points where this facility really suffered in the past.

Fault Case No2: Cold Room Thermal Insulation Failure fault detection and isolation utilizing only temperature measurements

Figure 2: Cold room behaviour due to thermal Insulation failure from the scope of temperature measurements.

Right after the FrigoBoss system’s commissioning and start-up, its automatic fault detection and identification functions picked up several faults throughout the facility. One of the failures picked up right away was the significant failure of the thermal insulation of Room 1. This was achieved by the automatic analysis of the time constants of Room 1. The successful detection of this fault right after the installation of the FrigoBoss system enabled the owner to decide not to place goods in this Room until the rectification of this insulation problem. Figure 2 above shows the behaviour of Room 1 (the one with the faulty thermal insulation) compared to Room 2 (with good insulation).

It can be clearly seen that during the occurrence of the fault described in the above section (refrigerant leakage) that the temperature rise rate of Room 1 with the faulty insulation is very fast and climbs above safe HACCP limits. Fortunately, this cold room was empty due to the decision taken upon the insulation failure alarm days ago. Thus, at the time of the refrigerant leakage fault, no goods were exposed to temperatures above safety limits, avoiding any health hazard and further increasing the money-saving potential.

Fault Case No3: Refrigeration Control System Inefficiencies detection

At this point, it is worth to be noted that the FrigoBoss system operates completely independent from the refrigeration system’s control system. In this specific facility, the refrigeration control system is a simple, old control system where practically each of its control actions has a local scope without considering the operational conditions throughout the whole installation. This naturally causes many inefficiencies with the most notable ones being huge energy losses and, in some cases, like the one we will examine here, even aggravates an already faulty condition.

The defrost operations of the control system are pointed out in Figure 3 below.

Figure 3: Detection of inefficiencies and caveats of the pre-existing, old control system of the refrigeration system.

The defrost operation is activated and controlled only by a simple timer relay and executes at predefined intervals, not taking into account any other operational condition of the refrigeration system. Thus, the defrost takes place, even if it is not necessary at all, as in this case. Furthermore, in this case, the control action leads to significant unnecessary energy consumption, as well as to the introduction of unwanted thermal load in the cold rooms. Furthermore, for this specific facility, where the power supply failures were common, each power loss resulted, among other unwanted effects, to reset the timer. This effect added to the significant inconsistencies of the defrost control action.

The FrigoBoss system greatly assists to identify and spot inefficiencies and caveats of the control system. We have used the FrigoBoss analytics to provide the customer with a detailed technical report on the facility’s existing control system inefficiencies and caveats, in order the customer to make informed decisions and act upon these insights. Furthermore, we were able to quote the customer for a completely new retrofit of the control system, based on the insights provided by FrigoBoss analytics, focusing on energy efficiency, safety and effective, integrated control of the refrigeration system as a whole.

Conclusions

Summarizing, this case study highlights the following advanced features and capabilities of FrigoBoss, apart from its standard function, which is the regulatory compliance to HACCP requirements and EN 12830:2018:

  • Achievement of incipient fault detection, fault isolation and fault prediction of various faults and failures, of the refrigeration system utilizing only the automatic analysis and processing of the measurements of an optimal number of temperature sensors.
    • The cases of the refrigerant leakage and cold room thermal insulation failure were used as representative examples to showcase these capabilities and features.
  • Identification of inefficiencies and caveats of the existing refrigeration control system and informed actions based on the insights provided by the FrigoBoss analytics.
    • The case of the ineffective and energy inefficient defrost control action of the old, existing control system of the facility, identified by FrigoBoss was presented in this case study.

Posted by Ioannis Bardakis CEeng MIMechE | August 26, 2019.