Anybody got experience with online oil sensors? These are meant to have continuous monitoring of oil conditions especially for big machineries.

Anybody have used these oil online sensors?

http://www.kittiwake.com/Default.aspx/ProductSection/10...SubSubSubSection/188

Why no one reply? Are online oil sensors new ideas compared to vibration sensors?

At our plant, there are a series of plant chemistry parameters that are continuously monitored by sensors and tracked by a dedicated computer system. We have never extended that to lubricating oil.

These may be good products. I'll play devils' advocate and give some reasons (excuses?) why this doesn't seem like a great idea to me yet. But I'll be interested to hear if I'm mistaken or there is more to the story that I'm missing.

These products are not yet widely used. Our approach is generally to wait for other people to use and prove a new product, rather than being the first to try new products.

Oil analysis can have a wide range of parameters: Moisture, elemental analysis, wear particles, viscosityetc. That suggests a wide range of sensors necessary to accomplish what can be done by a lab on one sample. (cost). I didn't see viscosity listed in the list of sensors. That is one of the most basic measures, and it might be argued that this is the only one that might demand immediate action... all the others are associated with degradation over time.

Another thing you mention this would be good for critical machines where frequent monitoring is justified. But for other machines, you'll still need the manually-sampled method with samples sent to the lab and results tracked and trended in a database.

So unless you convert all machines at the plant (pretty expensive I'd imagine), you'll have two databases to keep up with: one for your continuously-sampled machine and one for your normal machines. You'll still have to keep up the samping skill, purchase order with lab, etc. The only thing you gain is reduction in manpower for sampling some of your machines and more frequent (continuous?) monitoring of these parameters. So my gut says the cost will be higher. Since again these parameters (excluding possibly viscosity) are primarily long-term trended parameters which don't demand immediate action, it doesn't seem like there is a lot of benefit to justify the cost increase.

Josh,

Online oil sensors have been around for the last few years, but there has been this trade-off in precision, property capabilities, and cost. Online oil sensors generally fall into two categories: general oil condition sensors (such as dielectric type), and specific sensors for certain lube properties or contaminant levels. The general oil condition sensors have been around the longest, and are meant to indicate when the oil properties are changing, and should be sampled for detailed analysis. They will look for things such as moisture contamination, wear metals, and in some cases oxidation. Each of those can affect dielectric, and it will indicate a trend change. But it does not provide quantitative values about any of those properties.

On the other hand, certain sensors are out there for particulate counting, viscosity, wear debris, and moisture. Most of the affordable moisture sensors report relative humidity from 0-100%. That means if the oil has free water or is emulsified, it will read a steady 100%, so you can trend under those conditions. Also, RH% is not easily related to common testing for moisture like ppm or %. Particle counters are getting better, but most are very flow dependent in terms of accuracy. There are a few metal debris sensors out there, and they seem to keep getting better in terms of how small a concentration of debris they can see. Ferrous debris sensors are much more sensitive than non-ferrous metal sensors, and can typically see all particle sizes, while the non-ferrous are generally particle size limited.

Some magazines have run pretty good articles recently on the choices that are out there, and the recent trade shows have seen an increasing presence of vendors with various sensors. We have done some research to integrate sensors into filtration skids to improve the filtration process, and you can see papers on this topic at ASNE Oil Sensor Paper

Rich Wurzbach
rwurzbach@mrgcorp.com
Maintenance Reliability Group, LLC

What will the sensors monitor?

Four Most Relevant Parameters:

Oil Condition: TBN/TAN/Insolubles/Glycol/Water

Moisture: Up to saturation point

Ferrous Debris: All particles irrespective of size

Non Ferrous Debris: Individual particles >150micron

you monitor flow! and perform oil analyses. on some critical processes we monitir dailey and correct dailey as necessary.

Concern no. 1 - The products are not yet widely used. Yes, many plants do not want to be a guinea pig but the products were tested before commercialiation.

I heard US Army test about 1,000,000 used oil samples per year but out of this, 80 to 90 % are within specs. So if we put online sensors, there would be savings due to avoidable waste.

Yes, viscosity is not listed but their onsite oil test kit for viscosity can easily be performed to get its trending over time.

Other tests include water, insolubles, TAN & TBN. All can be done within 30 minutes by a mechanic.

In our case, lab analysis is reserved for advanced analysis eg identifying & sizing elements etc to know where are them from.

Also in offshore, it may take 1 month for the oil analysis results to travel from ofshore to lab and then return to offshore.

This message has been edited. Last edited by: Josh,

About 2 databases, these may not exist because the online sensor results can be connected to a PC or existing database I think.

Even then, don't you have 2 databases for vibration (one from DCS & another from portable analyzer) and can't you donwload results from vibration results from its analyzer to exiting database if we wish to combine them?

This message has been edited. Last edited by: Josh,

Definitions of the word "monitor" include:

To keep track of systematically with a view to collecting information.

To test or sample, especially on a regular or ongoing basis.

Sam, using the word "monitor" to describe what the sensors do seems like the right choice in words.

Electricpete, what do you think now? Convinced or not?

In many cases, you would either have two databases, or you would have extra effort to combine the results, as Josh suggests. I believe in time you will begin to see this change. Our laboratory uses the same software (Webview2000) to present the online sensor data in real time, and report results to customers on submitted bottle samples, so there is only one database. I'm sure eventually you will see others in the industry adopt this approach as well, as the sensor cost continues to drop and they become more prevalent. In the meantime, electricpete is correct in guessing that it is cost prohibitive now to try to place sensors everywhere. However, an effective oil sampling program plantwide, coupled with a targeted on-line oil sensor installation program on critical equipment, equipment with very short P-F curve failures with indications in the oil, and high-cost capital equipment makes a lot of sense and can be very practical and cost-effective.

Rich Wurzbach
MRG Power Labs
rwurzbach@mrgcorp.com