Transmitter on Tank

The idea is to have the proper control for the application, there are several types...

http://www2.emersonprocess.com/en-US/brands/rosemount/Level/Pages/index.aspx

If I understand correctly, the existing HH Level Transmitter (that trips on high high level for over flow protection) is being replaced with a level sensor based on pressure transmitter mounted at the bottom of the tank.

You mentioned the process fluid was oil/water. Are these in an emaulsion or suspended somehow, or are they as seperate phases? How is the composition of this fluid controlled? How stable is the composition of this fluid (will it vary)?

As you eluded to the differing densities would cause fluctuations in the measurement, if the composition of the fluid (ratio of water to oil) did vary significantly. As the pressure sensor measures the weight of the column of fluid (water/oil) above it. If the composition changed, then the differences in the density would change the weight of the column, and therefore the pressure reading would be different.

Unless the composition of the fluid was able to tightly controlled and therefore the average density was constant, this seems like a poor process choice (from what I understand).

Recommend alternative style LT (to use the same as the existing seems like the logical choice).

Hope this helps.

Anthony

At my last place of employment we measured lots of oil and water interfaces. We mostly used guided wave radars or displacers. Not sure what your application is.

http://www2.emersonprocess.com/en-US/brands/fisher/FieldInstrumentation/levelsensors/cageddisplacer/Pages/cageddisplacer.aspx

http://www2.emersonprocess.com/en-US/brands/rosemount/Level/Guided-Wave-Radar/5300-Series/Pages/index.aspx

I agree, by using a DP type TX on a tank containing varying medium SG's you will get a very compromised level protection. You can set it up as mentioned in this string to protect with a large span for error, but it really is a poor means of process control, and if your tank is very large you could be looking at tonnes of oil/water level discrepancy. Best off to go the extra mile and use guided wave radar. A good GWR will have settings for false diffractions, particulate diffraction and minimal wave dispersal and is probably the best means of level measurement in this scenario. However it's not the ONLY suitable method. If your using rosemount gear already. Talk to Emerson and they will be happy to sell you any number of level gauging equipment. Oh and if it's not a closed top vessel and you absolutely must use the DP tx, look into a nitrogen "bubbler" setup with a pneumastat. That way you can even get density measurement with your level feedback. I hope this helps.

My understanding of HH is to sound an alarm if other controls fail.With this in mind,knowing that the absolute value of the liquid in the tank is not very critical, as this should be taken care of by other devices the HH is simply an alarm to tell when any liquid reaches it.With the limited purpose in mind, and utilizing parts on hand, I made my recommendation.

A magnetic float ball with hermetically sealed reed switch would also work just as well for alarm purposes.Very rugged and reliable.

If you need an actual measurement, then sonar, or radar is the ticket in mixed liquids.

A bubble tube is also very good if a single liquid is being used, and nitrogen is not required if not subjected to freezing temperatures, and a glycerin bubbler can be installed in line to capture moisture, even when you think your instrument air is dry.

I have been out of the field for many years, and I am sure my knowledge has been superseded by newer technology, so I yield to innovation.

Absolutely agree. With one exception. Generally the High set point coencides with an alarm and the HH set point has a trip function. Turn off a pump, shut off an inlet valve, open a drain valve or bypass etc. It's the final safeguard and so if anything it should come under the closest scrutiny. I appreciate the input though.

As you have probably also seen, the HH can be used for many functions including changing tone of alarm to warble or pulse,etc.Many options are available from the alarm (relay) card, including motor control,solenoid or valve operation.Insofar as bullet proof, reliable and durable, the Gems magnetic float switch used to be top of the heap.The simpler the device, the less possibility for failure. No solid state devices to get wiped out by voltage transients, no tin whiskers to short out circuit boards, etc.

Redundancy with these switches is relatively cheap to implement at probably 10% of the cost of other methods.IMHO

Again, your perfectly correct. If the apparatus selected (whatever it may be) is purely for HH trip and or Alarm, then the simpler the better. The only reason I lean toward an analogue device with a high degree of graduation as opposed to a point detection device such as a magnetic float switch, is because of the control versatility it offers. Should control philosophy or set points change in the future. You may require a new connection flange rather than a simple software change.

Yes - like that. Use a DP transmitter (with measurement points spread as much as possible vertically) to measure the density of the bulk solution (would need to be as high as possible to be accurate, but must always be submerged). Then correct for the level measurement from the PT.

Innovative (and complicated but it's cool).

How tall is the tank? Why not run a long nipple or tube up to the HH level of the tank from the available fitting on the low port, and calibrate the transmitter for the lightest specific gravity fluid to say, 6 inches from full, or whatever the alarm level should be.Most oils are around .8sg so the worst case would be off by only about 20 percent of the calibrated range, so it could be a very narrow span and give fairly good control.

Example:Tank is 20 feet tall.HH level is 19 feet, 6 inches.Mount transmitter at 19 feet height, and calibrate it for 4.8 inches of W.C...If tank is full of oil,it will alarm at 19ft6 inches. If full of water, it will alarm at 19 feet 4.8 inches.You could adjust span or height of transmitter to get precise trigger point if more precision is needed;(a lower span will give more precision.)Or change alarm trip point.

These are off the top of head estimates, and may be off somewhat.

Good luck!

Not so sure about your estimates, but you make a very good point (GA from me).

Becaue the PT is looking at weight of fluid in the column above it. A 20% increase in SG would trip the PT, at a volume (and therfore a height) of 20% less. So if trip point was 20 ft, lower SG fluid trip would be at 20ft-20% = 16 ft.

But that said, if the OP is able to accept a 20% reduction in tank volume (to accomodate the difference in volume resulting from the max/min SG of the process fluids), then the instrument selected is a good selection.

We don't know if this has already been allowed for in the process control. Nor do we know the max/min compositions of the process fluid. If the fluid were to only vary in composition by 10% (oil / water). Then the SG may only vary by 2% (10% of 20%), and therfore the diference in height is much lower (on the 20 ft tank, only 4-5"s).

In any case using a PT as level control (in my experience) is a very reliable method. While radar is obviously more accurate (in this scenario), it has it's own drawbacks (condensation, foam, mist causing incorrect or false measurment). So there are good reasons to select against the radar type measurement - I know I don't like using them (for the reasons given).

The float switches are an excellent solution, but in many organisations these are also not used. Currently working on a job for major O&G company. Specifications forbid the use of digital (on/off) switches (not trusted for reliability). All measurement has to be done by analog instruement, with PLC singaling trip points. But this is extreme example, in my experience float switches are excellent and very reiable solution for level detection (and believe this is the best solution in this case).

Hope this helps,

Anthony

drawbacks (condensation, foam, mist causing incorrect or false measurement).

While I agree that these conditions all pose challenges for radar. When configured correctly with correct process conditions entered and proper false eco suppression they do work very well. The only time I have ran into trouble with them are on molten sulfur tanks when the head has become completely coated. That being said from his description it dosn't shound like he would have anywear to put one.

In my post, I indicated mounting the DP at 19 feet height,using a tube or nipple, therefor the only pressure measured is from 19 feet and above(use a pigtail to ensure that the DP stays full of liquid at all times). This would be a maximum pressure of 6 inches span to reach HH point.20% of 6 inches =1.2 inches, worse case, with oil on top of water, even less if it is in suspension.The alarm, or desired functions (output relay card) could be fine tuned by adjusting the trip points on the card.

IMHO( only one of 7 Gigi-opinions)

Gotcha - understand. If you position the DP just below the overflow, minimising the head of fluid, this makes the density difference insignificant.

You're right - this would work OK.