Cost of Steam Trap Failure - Remote Monitoring

Big-Al_2

Where I work, we have a large (150 HP) 15 PSI boiler that supplies steam to process equipment and a lot of overhead unit heaters. We have a number of steam traps that are difficult to access for testing. We're looking at installing a Spirax/Sarco remote steam trap monitoring system on them and need to see if it's cost justified. We're seeing about a 15% to 20% failure rate per year on these traps. I'm wondering what a failed trap would cost per year. If I'm reading my charts right, a 1/4" orifice would leak about 40 pounds of steam per hour at 15 PSI . . . for an annual cost of about $3,500. Does this sound right?

JUGHNE

I can't crunch the numbers for the cost of the wasted steam, but if it is anywhere close to that much it would be wise IMO to invest in a FLIR camera gun.

My last tool/toy was a FLIR TG 165....less than $400.

You can look at a ceiling CI rad and "see" the progress of the steam travel from 10' away easily.
Look at CI floor rads and see the difference between the inlet and outlet of traps.
Wall down the hallway of a nursing home and see what BB convectors are hot without entering the room.

This must be the most basic gun they have and if I were in your situation would consider something as such for serious trap checking. (10% investment for your estimate above)

EBEBRATT-Ed

Some say that traps should just be rebuilt every 3 years or so rather than spend time and money on testing.

especially the ones that are difficult to access, change them during a scheduled shutdown and be done with it.

Sailah

I have an Excel spreadsheet I made up for a client I can send to you. You would input your yearly run time, trap population by orifice size, and your cost for MM BTUs ( I was using $15).

For your 1/4" trap, 3500 hrs, $15/million BTUs I get $2685/yr in lost steam. or 54 lbs/hr.

We are in beta testing for our new clean sheet steam trap monitoring solution. We are leveraging the very latest in ultra low power, long range radio transmitters to enable a sensor network previously unavailable to anyone. So instead of relying on ZigBee, Wifi etc we are using LoRaWAN technology which has range in the 1-3 miles urban, 50+ miles line of sight. Can run for a decade on a single AA size lithium cell. Attaches to any piping with a simple split clamp and screwdriver.

I don't know how big your plant is but with a single gateway at my factory I cover the entire factory (steel building) and additionally about 1/2 mile radius around my plant in heavy wooden environment. I have a field trial at a large university in Boston and I'm canvassing the entire campus with a single gateway which can handle 100,000 sensors. The beauty of a scalable system like ours, is steam traps are just the beginning. Any sensor you can imagine could easily be networked with our platform. Light, humidity, temperature, tank monitors, smoke detectors, proximity sensors, electrical, machines running, batch done, CNC faults you name it. Can all be done. We obviously are first working on the steam trap monitor but a universal sensor box that will accept any of those sensors will be done soon.

The downside to LoRaWAN is the payload of the transmission is small. You can't stream video or monitor 100% of the time. This is due to bandwidth and battery life. I'm sampling traps every minute and transmitting every 5 minutes via packet. So assuming what you needed to monitor could fall within that basic framework of a general alert vs "FIRE ALARM" it is all possible and happening pretty soon. We should have a pretty good handle on our field testing by early next year. Working on approval by the utility companies here in MA to be eligible for energy efficiency incentives too.

Here's a sneak peak

Monitor on a 3/4" failed F&T at 5 PSIG



Monitor about the size of a deck of cards, but we could make it much smaller. I needed a commercially available housing that was steam tight so this is what I could get.



Here's the raw data being graphed, this was this morning. Anything above 1000 is failed, this trap was really failed. Where it drops to steady state is overnight when my boiler shuts down.



LoRaWAN architecture. This is the future of IoT technology. Everything from parking meters to street lights to your refrigerator will have this in the next 10 years. Maybe not @ChrisJ monitor top though...



LoRa radios also geolocate without GPS so asset tracking or just clicking a map view can bring up what traps are failed. This is at Northeastern University. My gateway is top right of the picture above NE-92, that campus is 1/2 mile in diameter, wish my wifi went that far.



Anyways sorry about the long winded reply but I'm about ready to discuss the monitor after a years worth of work and all the patent stuff is in place.

Sailah

Any yes I agree entirely with @JUGHNE that a FLIR & Accutrak VPE-1000 would be well worth the investment. But as you noted, in areas where it is difficult to access, those traps get ignored. And ignoring a bad trap is expensive hence the reason you are looking at a monitor.

PMJ

Big-Al said:

Where I work, we have a large (150 HP) 15 PSI boiler that supplies steam to process equipment and a lot of overhead unit heaters. We have a number of steam traps that are difficult to access for testing. We're looking at installing a Spirax/Sarco remote steam trap monitoring system on them and need to see if it's cost justified. We're seeing about a 15% to 20% failure rate per year on these traps. I'm wondering what a failed trap would cost per year. If I'm reading my charts right, a 1/4" orifice would leak about 40 pounds of steam per hour at 15 PSI . . . for an annual cost of about $3,500. Does this sound right?

I run a large boiler at low pressure(12) to heat tanks on a plating line. There is a pressure drop in all the coils so in your calculations I wouldn't use full boiler pressure as what a bad trap would be seeing. What gets by a bad trap in my line is nothing like 12psi. Also, when the tank is at temperature (which is a lot of the time) the supply valve is closed so a bad trap leaks nothing during all that time. Bad traps do waste steam but I'd be careful about the numbers depending on the application.

Also, no system is without a maintenance cost - even an electronic monitoring one. Figure something in for that. You would be paying to install two devices everywhere that currently only has one would you not? Every device has some expected life and random failure rate. May easily work out to be worth it - I'm not saying it won't. I'm just saying be careful about the numbers.

Big-Al_2

Thanks, everyone. Good information for sure. The insight into LoRaWAN is especially interesting. I'd like to get more utility and power consumption monitoring online. Sounds like the way of the future for remote monitoring without running cables everywhere.

Sailah

Here's my first 100 getting installed tomorrow

The dashboard is located at www.steamiq.com

It's still a rough draft

Koan

that is so cool!

Sailah

Big day today, I installed about 50 monitors at Northeastern University. We are getting fantastic results. We checked every trap with an ultrasonic tester prior to installing a monitor to validate what we are seeing.

Sitting at home on my iPad I can tell what is leaking and what is good. Unfortunately some of the leaking traps there were Barnes & Jones lol. But all the manufacturers were represented ha.

The system needs a couple days to settle down and I need to input my notes from today. Just driving the monitors around in my truck caused them to get a jolt and register a fault. But the ones that are truly leaking are easy to see as they are registering fault levels of 300,000+. Average steady state is 30k.

In one area we installed about 10 monitors in a very large rooftop coil with a preheat. It was like a walkin wind tunnel, 30mph winds in there with the fans going. Monitors easily picked up the bad ones.

Pretty exciting stuff for a steam trap geek.

You can see right when I installed this one on a bad 1 1/2" FT trap in the air coil before lunch. Just think of the steam being wasted 24/7 on that...

Sailah

This is inside the coil.

The hoffman top left is bad NE-87 and the big FT30 you can't really see under the coil is bad too. Really bad. Same type of trap you can see in the foreground. 1.5" pipe...



Bad one top middle NE 91 on Steamiq.com

neilc

so where does one sign up for this ?

Sailah

I'm not yet ready for prime time. Working on patents, business model etc. hoping to be ready to go live this summer.

This trial I installed today is going to be a long term test site 6 months.

Another interesting thing I found was the range on the devices. Was getting an incredible range out of some of them through buildings, underground, steel enclosures.

At northeastern their campus is about 1/2 mile in diameter. I am covering the entire campus, and probably half of south Boston, with a single rooftop gateway about the size of a home router. I placed a couple smaller gateways in basements of large buildings to fill out the coverage.

Neil I'm happy to talk to you about your specific application if your interested. I have a limited number of monitors I had produced, mainly due to the radio chip undergoing initial production. These are also prototypes and I chose the housing as an off the shelf NEMA, watertight housing vs getting an injection mold done. Once we validate our testing, I'm going to redesign them much smaller and have a custom case made injection molded etc. these are about the size of an older iPhone, but 2" thick. I could probably make them 1/4 the size.

Battery life is projected to be 7 years with current reporting times every 30 minutes. It samples every 1 minute and reports to the dashboard every 30 minutes. If you stretch those out you can really extend the battery. Reality is I probably only need to sample every hour and report once a day for the production version but these I want instant results for the nice folks that are hosting this.

Sailah

So far testing has been excellent. The facility I'm testing in has 3000 traps so in some sense I'm only getting a small picture of what's going on. But we've found, and fixed, a lot of problems so far. Found a trap that had lost its float and was blowing wide open through a 1/2" hole @12psi. Found scores more with failed air vents. Found a few that were getting back pressure from a much larger trap and were failing in reverse.

I have about 25 monitors left and looking for a smaller test site to run concurrently to Northeastern.

If you have a small plant or large house with around that many traps and are willing to give it a whirl I'd like to discuss and see if it's a fit. If you are in new England I'll come and install myself, if not it's a pretty simple process.

The monitors are battery powered and attach to the outlet using a split pipe clamp. Installation takes a minute. They communicate via a gateway that plugs in to an outlet and uses an ethernet port to transmit back to our network server. Range in a house is about 1 mile. My plant is 50k sq ft and I easily cover it and about half the business Park. The gateway would then communicate via the dashboard at steamiq.com. I'll set up a section just for your install.

So what's in it for you? You'll gain access to what you can't see in your system. If you suspect you have a failure it'll tell you. I'll send you the repair parts no charge. The goal is a system running at peak efficency.

I'd also need your permission to use the results for our testing study. Will probably run for remainder of heating season.

Let me know if you are interested and we can discuss further. Thanks

neilc

well I could probably take you up on about 10 of them,
low pressure residential,
mainline traps only, we're not talking radiators, right?
Vanderbuilt home(1937) converted to private school(1960),
Westport Ct
wanna discuss ?

Sailah

I can do radiators as easily as F&T traps, buckets, TD traps. If it's a steam trap I can monitor it. As long as we can clamp the split collar on the outlet I'm good. Could you make use of all 25 doing a mix of rads and FT traps?

Yes happy to discuss give me a jingle Tuesday. Thx

Koan

@Sailah
don't know if this is of any interest to you.
I have 11 traps on 3 floors (all with B&J guts). Residential Hoffman two pipe Vapor system. Returns out of the traps are 1/2" pipe. How big (diam and length) are the split collars? Let me know if you want to consider?

Sailah

Kelly,

I appreciate it but I think at this point I have more interest than I have monitors unfortunately. If something comes up I'll be sure to reach out. Thanks! Peter

Koan

Peter,
no worries - mine is probably not the ideal application.
Out of curiosity, would you comment on how the clamp senses the trap failing? I assume it is either through high (blocked open) or low (blocked shut) temperature? What temps do you use as thresholds in case I test manually?

Also I was away over the holiday, so I did not get a chance to send back the 2CR in exchange for the Big mouth 3BM. I was going to polish the 2CR before sending it back , but I'm having trouble scheduling with returning form vacation!
Kelly

Sailah

Hi Kelly,

I'd be happy to comment, right after the patent application is submitted lol. No we don't use temperature for monitoring other than the ARM chip on the board has a built in thermistor. The reason that temperature isn't very accurate is that other failed traps can mask the functionality of a good trap by feeding steam up the outlet pipe. It's not very common that that happens but enough that we don't consider temperature accurate enough to rely on.

For the home user it is fine. Typical inlet temps for low pressure systems are 213-220F. You'd be looking for a 10-30 degree delta at the outlet. So inlet of 215, outlet of 200 I would consider functional. If inlet is 215 and outlet is 209, I would call that trap suspect because even though it's below saturated steam temp the drop probably has a lot to do with the piping vs what's inside of it.

I just picked up one of these to keep in my FLIR case. It's not the most accurate ultrasonic sensor but it's certainly good enough for casual use. I've used mine for a bunch of things around the house too. Pretty decent deal if you ask me, new they are far more. This particular model uses an airborne probe vs the more expensive ones that use structure borne ultrasonic but I find it's still pretty accurate.

http://www.ebay.com/itm/AccuTrak-vpe-Ultrasonic-Leak-Detector-Tester-Basic-Kit-/172470404278?&_trksid=p2056016.m2518.l4276

Eastman

How does LoRaWAN geolocate?

Sailah

There are GPS Chips in the module as well as triangulation from the gateways. It's used in asset tracking, vehicle tracking as the battery consumption is so low.

Eastman

Does the GPS drain the battery? Do you normally turn that feature off and rely on just gateway triangulation?

Sailah

Sorry I was incorrect on the GPS. As to battery life we are getting 7-10 years from a single AA lithium battery as the device has a super low sleep current, 2 ųA.

https://www.semtech.com/Press-Releases/2016/Semtech-Announces-Availability-of-Next-Generation-LoRa-Gateway-Reference-Design-Platform.html

Eastman

How accurate is the triangulation from the gateways?

Sailah

I'll ask my network provider they are lora experts

WayneMech

Definitely keeping an eye on this. The ROI is beyond question, for anything beyond residential. I would love to be in on this as early as possible.
-Dale Greenhalgh

Sailah

I will keep things updated.

Don't discount residential. I think I can do it based on where you live. What I mean by that is if you are within range of a public Lora gateway, and there are many, I just eliminated the gateway cost. Further depending on your utility there may be incentives for efficiency. I've submitted for MTAC approval here in MA I assume others have a similar technology vetting board.

Does anyone here know about Enertrac?

http://www.enertrac.com

It's a tank monitoring device for both propane and heating oil. The technology they utilize to help home heating oil delivery drivers optimize their routing is pretty cool. The technology is the same for the radios. And since they have the gateways currently deployed on existing cell towers I can utilize those to punch out to residences. Large commercial buildings have such a high node density it just make sense to bring the gateways to them. The long range low power is perfect for sensor data harvesting. The range is pretty incredible too. In ideal conditions you can get 50-100 miles range. Enertrac told me they have tank monitors on Nantucket hitting towers on long island!

BobC

i have an old book of limericks, you would be amazed at how often Nantucket gets mentioned therein.

All joking aside I think the work you are doing is groundbreaking, that information can save a lot of energy.

Bob