chemical calibration - confused

have something that i am confused about and hoping someone could get me going in the right direction.
Have inherited a chemical distribution system that has a slc500 plc and it is used to inject a certain amount of chemicals into a tank, depending upon what the operator enters.
the chemicals go through a flowmeter that gives 1 pulse for every 1/5 of a gallon.
operator will select for 1-50 gallons, depending upon chemical.

what i am wondering is what would be the best way to set up a routine to double check and calibrate to verify accuracy?

for example lets say you need to add 10 gallons, so after 50 pulse you would have the 50 gallons, but what if after the 50 pulses you only had 49.5 or 50.25 gallons. Confused on the best way to handle this.
I know we have some other chemical systems that are controlled by pc equipment, and when you calibrate them, you select to run a certain amount, for example 5 gallons. with these systems you would run the 5 gallons into a know bucket that is premarked to show the gallons in increments of 1/10 of a gallon. so after you select the 5 gallons, you check to see how much was actually put in the bucket, then there is a section in the menu on the screen to put in the actual gallons that was put in, it then goes in and calibrates this amount to be the 5 gallons.
Any suggestion on how to do this with a slc500?
sorta got me confused at the moment, so any advice would be greatly appreciated. Its just one of those things that for some reason, has me confused, ever have one of those? I know it must be a lot simpler than i am trying to make it in my mind, but just kicking my butt at the moment.
Thanks

This doesn't exactly answer you question as far as how to best do this in a SLC500, but I think it needs to be answered first.

What is your allowable volume error for a 1 gallon delivery?

For me, first knowing the exact way the pulse is designed, and how the PLC is to count and tally pulses would be the beginning.

Secondly, for a system using 1-50 gallons (or 5 to 250 pulses), if the pulse ON and OFF signals are of equal duration, I would think you couldn't expect better than a +/- 10% accuracy at 1 gallon (the % of 1 pulse change for your total # of pulse changes, or one on ten). For 50 gallons, you couldn't expect better than +/- 0.4% (or one in 250), which is only 0.2 gallons.

I would think that for a desired delivery amount of 1 gallon on the low end, a 5 pulse per gallon input leaves a rather large percent of error, depending on how the pulse is created, and how the PLC treats the pulse inputs.

Is the ON time of each pulse equal to the OFF time of each pulse?

How does the PLC count - that is, what does it use as a pulse count in order to add that volume to the batch total? Does it count from the start of a pulse to the start of the next pulse as 1 count? Or does it check to see if the pulse is ON or OFF at the start of the delivery and count 5 of those statuses per gallon, depending on what was present when the delivery started?

And when does the PLC stop the injection? Does it try to stop the injection at the start of the pulse that equals the desired volume? Or does it wait until the final pulse is over in order to start a new batch without the pulse input present?

How does the PLC treat an input pulse that remains at the end of batch one to the beginning of batch two?

Is there a delay from the PLC output turning off to the process stopping delivery of the chemical, or is it rather instant due to a solenoid?

Of course, these questions are less important the higher the desired volume to be added, because the error of one pulse becomes less and less as the total number of pulses goes up.

What is causing you to think there is the need to do a calibration?
Does your QC process do a check and what are their results?
Is your QC doing the right check in the right way?
Maybe "if it aint broke dont fix it" ???

OK now for your question
How to get the calibration data results into the PLC?
FIRST since you are trying to ensure you are delivering the correct volume then your idea of a calibrated volume (graduated bucket) is perfect.
SECOND if your calibration is withing tolerance then you are OK and need to do nothing
THIRD if your calibration shows an error then I do not think I would bother with the PLC. I would check the delivery equipment to ensure if it is operating properly. Most "PLC problems" I have seen are either a sensor or an operator not the PLC itself. This would be doubly true if the system has operated correctly in the past with NO changes.

One more thought
could temperature change (thus density) be changing your results??

Dan Bentler

I agree with the others, in this type of system using a pulse meter you are not going to get much better results. I would think anything within 1% would be acceptable and pass specs.

You could always look into swapping that meter out with one that generates higher pulses.

If accuracy is that critical you may also want to look at a coreolis type meter.

Also there are other factors to consider such as head pressure and temperature.

Finally, what do you do with this chemical? Do you then mix it with water, other chemicals and hand adds.

I'm not sure what your original question was after....If it was how to increase accuracy then the above posters have done a fine job of explaining what you can do. If you are asking how to do a calibration check, we do them all the time with oil injectors and flow meters. Simply have a calibration mode where an operator enters a desired amount. Run the delivery system for the # pulses to achieve that volume and weigh it on an independent scale. If you want to do it automatically (periodically), you could have a large 200 gallon container sitting on a scale controlled by the PLC (remote tare and analog weight). Whenever the system is idle or in a state that's permissible, switch a valve, tare the scale, run a sample into the container and weigh it. Alarm on deviation. Of course, your alarm has to factor in the tolerance of your scale, your flow meter, etc. to keep from generating false positives.

Thanks for replys

Great ideas on how to increase accuracy.

The pulse meter is a Toshiba mag meter, the only thing i know at the moment, is that it gives 1 pulse for every 1/5 of a gallon.

I will actually get to see the system next week when i go to dominican republic. I agree that I do not think that 1/5 of a gallon for every pulse is going to be very accurate.

The things that I do know at the moment, is that it is a slc505 plc, will get a copy of the program next week.

What I am confused about is if for some reason, they want to do a calibration , not quite sure how to handle it in the slc505. I know that the output of the meter goes into a input on the plc, and if you need 2, or 3, or 5 gallons, then the pulses are counted until the desired pulses are seen. What i am unsure of is if for some reason you want for example 2 gallons, after you get the 10 pulses, and if for some reason you only have 1.5 or 1.75 gallons instead of the desired 2 gallons, then how would you handle that in the code. An example would be greatly appreciated.
The only thing I do know about the system is that a integrator in DR installed the system, it has been running for about 6 months now, and integrator is no loner availiable, and our QC folks in DR are wanting to know now if the system is delivering what it is suppose to. My understanding at the moment is that it is a bleach type chemical, and if it is not exactly what is asked for there is not a problem, but now a new manager has taken over the area and he wants to know if he asked for a certain amount of gallons and does not get it, then why not.
Any example of how someone may have done this in the past would be gratly appreciated.
Thanks

First, you need to insure that your meter is functioning property and then its only a matter of setting up a calibration number in the SLC. It would be best to give the supervisors the ability to change this number through the HMI.

Say for instance, you need 100 gallons, this would be equal to 500 pulses (setpoint) in the SLC. If you were constanstly only getting 95 gallons, you would need to increase your setpoint to the SLC.
100 / 95 = 1.05. 1.05 * 500 = 526. 1.05 would be your calibration number. In the SLC you would take the old setpoint of 500 and then multiply it by your new calibration number.

Like I said first, you need to insure that your meter is functioning correctly. There may also be some type of adjustment on the meter that may be able to compensate.

Irondesk,

Basically, you can't fix a meter problem in the PLC code. If the code is right, there is very minimal tweaking you can do with scaling the amount per pulse to exactly match the flowmeter. For example, if your meter is giving you one pulse for every 0.4 gal instead of 0.2 gal, you could adjust by multiplying by a 0.5 calibration factor but your accuracy is lowered and nothing you do in the PLC code will fix that.

There is usually a calibration factor in the flowmeter and a recommended procedure from the manufacturer for adjusting it based on testing. This is the right place to adjust your calibration. However, most current "smart" flowmeters have very good factory calibration and rarely need to be adjusted in the field. I generally do a "bucket check" to verify they are working correctly, but I very rarely change the cal factor.

So, do your test. If the meter is within tolerance, DO NOTHING to the meter, just document the results. Your equipment is probably not as accurate as the factory test rig so minor tweaks are not advised. If the meter is not within tolerance, find out what is wrong and fix it. It might be an installation issue or damage to the meter. Find the root cause and repair or replace.

Good luck,

Thanks

Will look over the program as soon as i get it.

Great info

irondesk40 said:

What I am confused about is if for some reason, they want to do a calibration , not quite sure how to handle it in the slc505. I know that the output of the meter goes into a input on the plc, and if you need 2, or 3, or 5 gallons, then the pulses are counted until the desired pulses are seen. What i am unsure of is if for some reason you want for example 2 gallons, after you get the 10 pulses, and if for some reason you only have 1.5 or 1.75 gallons instead of the desired 2 gallons, then how would you handle that in the code.

Click to expand...

Unless you have a means of independently measuring the flow, you can't do what you are asking. The PLC only knows the info it is given. If you tell it that a pulse is 1/5 of a gallon, then 10 pulses will EQUAL 2 gallons EVERYTIME. There is no such thing as the PLC saying oops you got 1.75 gallons. The only way to do this as I said before, is you have an independent measurement. Either by weight or redundant flow meters. Whether online or offline. Once you have this, then the code is easy. As an example:

Operator enters calibration volume (let's say 10 gallons)
System runs and collects 50 pulses and says that's 10 gallons
Operator measures on a scale and the scale reports 9.8 gallons
You calculate that each pulse was actually .196 gallons, not .2 gallons (9.8 / 10 * 2)
So, you adjust your pulse scale factor to .196 instead of .2.

I would probably insist on 5 weighings and take the average as you don't want one aberration to skew your scale factor. Even better, take 10 weighings, throw out the highest and lowest, average the remaining 8 and do the same math as above.

Thanks for info