This article provides simplified explanations to help gain a better understanding of the functionality and performance of a pH and ORP sensor. 

 Understanding Sensor Performance Rev1.3

The importance of understanding pH/ORP sensor lifetimes.

 
Monitoring Performance

For all plants and sites, it is important to understand how your sensors are behaving, both from process control and cost management perspective. pH and ORP sensors are by their very nature of operation a consumable product; that is to say, as electrochemical sensors they are both simultaneously poisoned and consumed by the process. The more chemicals or extreme conditions they are exposed to, the faster they will expire.  Once expired they must be replaced.  In tough applications, this can be a matter of weeks or months, while in milder applications it can be a matter of months or years.  Additionally, that can consume a significant amount of human resources as they need constant care and attention.

Regardless of the application, it is critically important to establish benchmarks for sensor lifetime and performance so that we can:

• Optimise maintenance to minimise the labour involved in sensor management (often the single biggest expense)
• Improve sensor lifetime and/or performance
• Quickly recognise when something has changed within our plant or operation and take corrective actions
• Have an adequate inventory for replacement sensors
• Make continuous improvements to our sensor measurement program (including design and instrumentation)

No History? Start Recording Now!

When a plant has no recorded history of its sensor experience it is impossible to assume or estimate how long a sensor might last or perform.  No amount of theoretical analysis can tell us how it will behave, nor can we translate the experience of one site to another.  The ONLY way to understand this is to take a scientific approach and properly monitor and record the behaviour of an installed sensor.  Thankfully, Smart Digital Sensors take a lot of the hard work out of this process by recording pertinent data such as the time in service, as well as other critical variables such as the max/min temperature the sensor might have been exposed to, as well as some diagnostic information on how the sensor most recently performed.

However, that is only part of the picture. While sensors are made to very fine manufacturing tolerances (and vary very little from the sensor to sensor), the same cannot be said about plant processes. Much of the variation that you will observe in your sensor life is in direct correlation to changes in your site’s operation and plant behaviour.  If you are going to get a proper understanding of sensor performance, we must allow for plant variables that will have a detrimental effect on sensor life. 

Considerations that affect the performance of a sensor:

• Location

  Each location in your plant will have a different impact on your sensor depending on the forces and chemicals that are present. You cannot treat all parts of the process equally, and each location should be assessed in isolation. Even if you move a sensor’s position within the same location it should be noted.

• Plant Stability
  How stable is the plant process?  New plants and start-ups can be highly unstable for months until all systems settle in. Even within well-established plants, production or process changes can occur intentionally or unexpectedly. This can include things such as: 

  • Optimise maintenance to minimise the labour involved in sensor management (often the single biggest expense)

  • Improve sensor lifetime and/or performance

  • Quickly recognise when something has changed within our plant or operation and take corrective actions

  • Have an adequate inventory for replacement sensors

  • Make continuous improvements to our sensor measurement program (including design and instrumentation)

           The above changes can expose the sensors to higher levels of chemicals, different contaminants, or stress that is otherwise not considered normal.

• Temperature
  Temperature and thermal shock will affect a variety of sensors in different ways. Are the sensors exposed to significant swings in temperature or are they operating near their limit?
  
  
• Start-up/shutdown
  The sensor can experience wild swings in conditions on startup and shutdown that can cause significant deterioration or failure if not managed correctly.
  
  
• Maintenance and Care
  This is arguably the most critical aspect that we can easily control. In most instances, poor sensor performance is directly linked to insufficient or improper maintenance techniques. Always consult with your representative on an appropriate and timely maintenance regime. No sensor is set and forgotten, and sensors require the correct cleaning agent to remove microscopic (or even visual) deposits carefully not to damage the sensing element. Failing to clean a sensor properly can result in the sensing element being poisoned at a much faster rate.
  
  
• Human
  There are a variety of human factors that affect sensor performance. Always note when there is a change in personnel or maintenance crews and see if the sensor performance changes (it usually does).  Not all people take the same amount of time and care when calibrating, cleaning, or even exchanging and servicing sensors. Some may even choose to discard sensors prematurely rather than having to clean them.

To get a clear picture of sensor performance and to enable a scientific approach to improve sensor management, it’s imperative to collect accurate information about what is happening to each sensor. Observing a single-use sensor in just one location is of little use from a statistical standpoint.  Without our direct knowledge, it could have suffered the ill effects of any of the above events. Instead, to build a clear picture of sensor performance, we need the history of at least 5 consecutive sensors from the same location, and ideally, 10 is considered the required number for an accurate statistical trend.  It is important to have as much data as possible and the following information is the minimum that should be recorded and collated so that we can develop a proper understanding of how the sensors are performing; relative to their environment and the maintenance and attention they are receiving.

Notes: 

• Calibration frequency should be established following the Turtle Tough Guideline for establishing calibration frequency.
• Where possible, cleaning should ALWAYS be performed before calibration
• Calibration Standards should ALWAYS be fresh
• Cross verification against lab or portable methods should be reviewed by authorised personnel. The significant error can result from poor cross-verification methods, resulting in premature rejection of the sensor
• Refer to Turtle Tough Guideline for Establishing Sensor Health
• You may wish to expand this table to include other important information, such as operator name etc
• If in doubt, return your sensors to Turtle Tough for a free evaluation
  

  ---

Making Sensor Improvements – A process of trial and error

As a custom builder, Turtle Tough can tweak your sensor design to optimise its performance.  Typically this cannot be done until we have sufficient information (like that tabled above), to make an informed decision about changing the configuration of your sensor. If you wish to review your sensor performance we are happy to assist with this process and possibly offer alternative designs provided that you:

• Can provide us with complete and accurate information (as above) and have tried at least 5 sensors in the same location.
• Return those sensors to us or your nearest representative for analysis
• Provide a visual image (picture) of the exact installation showing how the sensor is installed
• Understand this is a process of trial and error, and there are no guarantees, free samples or trials. We are simply, as an equipment vendor, offering an alternative configuration on a good faith basis that might better accommodate your conditions. Any subsequent recommendation we make is at your discretion whether you choose to purchase that product or not.  We make no guarantee as to the performance, and once again we stress that a single-use instance does not represent a statistical trend and that several sensors should be tried between making and further changes.