Preventive vs. Predictive

Posted on July 5th, 2018

Do You Use the Wrong HVAC Maintenance Technique?

Performing routine HVAC preventive maintenance (PM) can waste time and money and leave a facility vulnerable to critical failures. To avoid this situation, facilities should consider a mix of maintenance types to maximize reliability and efficiency while minimizing costs and inconvenience. The following describes four principal types of HVAC maintenance programs and when and how to use them.

Run-to-Failure (RTF) Maintenance

Run-to-Failure maintenance is by far the easiest type of maintenance when it comes to planning and implementation…don’t do anything, until something fails. RTF maintenance is perfect for numerous assets but catastrophic for others. The classic example of replacing a failed incandescent light bulb illustrates the pertinent criteria for RTF:

  • There is no secondary damage to other assets
  • The failure has no economic, safety or operational significance
  • The failure is obvious (not concealed)

The advantages of RTF maintenance include minimal planning and simple implementation; however, the disadvantages can include the unpredictability of timing for repair work and inventory costs. While minimal planning is attractive, the tracking of failures should be a priority in order to detect failure patterns and assess inventory levels.

Clearly, RTF is not a maintenance philosophy to be used for major, critical assets. Operating a chiller to the point of failure would probably be called “accelerated replacement” maintenance. Besides light bulbs, RTF maintenance can be applied to HVAC assets such as easily-replaceable filters, membranes, seals, switches, ballasts and monitors.

Preventive Maintenance (PM)

Preventive maintenance is maintenance performed regularly and triggered on the basis of time, usage or a specific metric or event. Changing oil in your car is an example of PM. You change the oil after a certain number of months (regardless of miles driven) or you change the oil after a certain number of miles are driven (regardless of time since the last oil change). You might even change oil after driving off-road (event specific).

PM requires scheduling and planning not needed with RTF maintenance, but this keeps operational assets running as designed, hopefully for their expected lifetime. Regular preventive maintenance can prevent failure for both critical and non-critical assets such as all types of mechanical HVAC equipment requiring lubrication, filters, flushing, cleaning, adjustment, tightening, maintaining fluid levels, etc.

The primary advantage of preventive maintenance is the simplicity of organization and execution. The marshalling of needed supplies, parts and resources and the scheduling of downtime to minimize costs, inconvenience and production loss. The disadvantage of PM is that time and money are wasted when it is performed too often or unnecessarily or it results in damage to the asset. Changing car engine oil at 3,000 miles or three months is a very conservative but costly and unnecessary preventive maintenance effort. Many current oil change recommendations reduce oil changes to 5,000 miles or six months to save time and money and reduce the potential for stripped threads, oil leaks or more serious maintenance-related accidents.

Condition-Based Maintenance (CBM)

Like preventive maintenance, the goal of condition-based maintenance (CBM) is to perform maintenance on critical and non-critical assets to avoid failure and extend asset life. However, rather than triggering maintenance on the assumed condition of an asset (based on time, usage, etc.) as PM does, CBM triggers maintenance based on the asset’s actual condition in order to further reduce maintenance frequency, cost and failures.

Determining an asset’s actual condition can be done continuously such as with remotely-monitored sensors, or it can be done manually at regular intervals by visual inspection, performance analysis or other types of tests. Asset condition tests include: vibration, electrical current, pressure, temperature, flow, acoustic, oil analysis, ultrasonic, etc. This would be similar to using real-time analysis of car engine oil to trigger an oil change.

CBM is much more complex to implement and manage than PM and generally requires some type of computerized maintenance management system (CMMS). The condition-monitoring hardware and software, setting up and analyzing of databases and staff training are an added cost of CBM compared to PM.

The cost and reliability advantages of condition-based maintenance increase with the value, complexity and criticality of the HVAC assets. Aircraft engines are good CBM candidates being valuable, complex and critical. This is why some aircraft engines are outfitted with as many as 5,000 sensors which are sampled every second the engine is running.

Implementing a melded RTF, PM and CBM maintenance program is best done with the help of HVAC service professionals capable of designing, installing and supporting automated and remote monitoring, data analysis and a computerized maintenance management system.

Predictive Maintenance (PdM)

PdM is the future of maintenance. It is closely associated with CBM and relies on many of the same sensors and tests. Like PM and CBM, predictive maintenance aims to reduce CBM maintenance intervals and costs. To do this, PdM puts an additional layer of information on top of that from CBM in order to define an asset’s condition and operation relative to its point of failure.

For example, a critical chiller bearing might be fitted with a 3-axis vibration sensor, an oil-line metallic wear debris detector and an infrared temperature scanner. Data from these devices provides information about changes in vibration levels, metallic chips detected and rising temperatures, but the data doesn’t tell us whether the bearing is about to fail or not.

PdM works to answer the failure question. To do this, data captured during the failure of similar bearings in similar applications or data from a variety of computer modeling techniques provides insight on the conditions exhibited by an asset prior to failure. As more sophisticated sensors and software are deployed for HVAC assets, the availability of PdM data and knowledge will continue to grow and make PdM easier and increasingly effective.

Predictive maintenance requires all of cost and effort put into CBM in addition to the costs of obtaining and learning to utilize PdM tools and information.

Developing the Right Maintenance Plan

As more HVAC assets include smart maintenance and monitoring features such as wireless communication, remote sensing and predictive software, automated maintenance programs will become easier and more common. It can’t be long before smart buildings have a single system that combines Building Management & Automation Systems (BMS & BAS), Energy Management Systems (EMS) and Computerized Maintenance Management Systems (CMMS) to simultaneously optimize comfort, reliability, energy efficiency and maintenance costs.