11 replies to this topic

[SPOOK]

Hello,

I am trying to design a vaccum system(water jet ejector) for removing vapors from an agitated wash water tank for PTFE.
The system is like this, 5 m3 of water at 90 deg C is filled in the agitated vessel of 7m3 volume. Now 450 kg of PTFE is added to the tank for washing. Now i want to remove what ever vapor us generated in the empty space above the liquid level in the tank. For this I have to design a vaccum system that will generate 740 mm Hg abs. pressure at the ejector suction nozle.

Now in order to calculate the non condensable load i.e., mainly air in my case I followed method suggested in book 'Applied Process Design' by Ludwig and I ended up with a very large value of air leakage 15 kg/hr.

The summary of air leakage is as follows:
flange lekage up to 6 inch - 0.5 lbs./hr
flange leakage above 6 inch - 0.8 lbs./hr
stuffing box leakage for agitator shaft - 1.5 lbs./hr
sight glass leakages - 1 lbs./hr

[b]The suggested method also asks me to multiply the calculated air leakage by a factor of 1.5 to 2.

My query is following this suggested method i end up with a very large size of the ejector for which fairly high flow rate of water is required. Also the fact is the required vaccum is just 20 mm Hg g. for which by the suggested method, in Ludwig, is not giving a cost effective water jet vaccum system.

Can any one suggest me any other guidelines for vaccum system design.

Your help is highly appreciated
spook

[DRS]

15 kg/hr doesn't seem to be very high for me. Since your vacuum is 20 mmHg (g), presuming the same air density at that of atm. pressure, the leak rate is 15/1.2 = 12.5 cu.mtr/hr or 7.35 cfm. Nevertheless, just check at what vacuum level the leak rates are mentioned. If you have a receiverr or catchall, you can check the actual leak rates of your system.

Interactive Engineering Fundamentals Handbook at http://www.sihi-pump.../Itemid,26.html is the best resource I ever found out (In fact, it is the only resource I saw in about 14 years that deal with design of vacuum systems and absolutely free of cost). You can see recommendation abut general leak rates in that document.

[SPOOK]

thank you DRS for your response.

Actually, I cannot measure the actual leakage obviously. Otherwise, I would have not put up the post. anyways, i went through the hand book it has only one graph the gives air lekage for the corresponding system volume. that graph is also mentioned in ludwig which i also have access to according to that graph the leakages comes to around 4 lbs/hr ie, 1.82 kg/hr. which is way less than my calculated value of 15 kg/hr.

[breizh]

Use the search button .you will find a workbook published by MR Montemayor called "producing and maintaining vacuum" .
Breizh

[Art Montemayor]

Spook:

I'll make an exception this time; but you should use your initiative and self-resources through the "SEARCH" feature in our Forums. If you need the information, you should work for it and searching is not that much work.

Go to: http://www.cheresour...x...

Look for Post #18 and you will find the downloadable Excel workbook dealing with producting and maintaining a vacuum.

[SPOOK]

Thanks for specifying exactly where would i find the resource Art. I have already gone through the spread sheet before you actually specified the path. I dont quite know why you felt that I have not gone through the spreadsheet. The method described in the spreadsheet is exactly the method given in Applied Process Design by Ludwig which I have followed by which a very overestimated value comes up.

I would be very grateful if you share your vast practical experience and guide me that the values that I have reached is appropriate. Also if there is any other comparative method which I am not aware of that I can use to cross verify my results.


Cheers....
Spook

[DRS]

Better be conservative on leakage rates. There will not be any issue if your evacuating system (for correctness I refrain from using the name vacuum generation system ) capacity is smaller when compared to the system volume as the time taken for evacuating the system, to a certain absolute pressure, increases but you can still acheive it.

In case of leakage, the system limitation comes into picture, at certain vacuum, where the acfm of leaking air equals capacity of vacuum pump (in acfm). Beyond this you can't further reduce pressure. This may not be a problem in your case as you are not going for deeper vacuum levels.

HEI has a graph for air leakage into vacuum systems at various pressures but it is in line with what is mentioned earlier and now. You can ask a manufacturer to take a trial on your system regarding leaks. Evacuate the system to 20 mmHg (g), isolate the pump and observe the pressure rise and time. You can assume the process as isothermal and do the leak rate calculation.

Just out of curiousity, why are you not thinking of a liquid ring pump?

[Art Montemayor]

Spook:

I can't give an opinion as to whether your calculation results are OK, valid, or incorrect without having the actual, detailed calculations in front of me. I am not questioning or challenging your method or your math; I am simply expounding conventional engineering methodology: all engineering calculations should be checked - preferably by peers.

I have used Ludwig's recommended methods and factors for many years. The relationships I included in the Workbook I prepared for training may resemble Ludwig's, but most of them are derived from other sources - just as authoritative and experienced as Ludwig. The empirical relationships I show have been used to calculate the vacuum requirements for some pretty big projects - and they worked well enough for the final, detailed engineering specification and design.

The design of a proper vacuum system can't be described in one or several postings. Articles and books have been written that fail to cover everything of importance in generating and maintaining a vacuum. In June of 2007, I wrote some very lengthy, detailed posts in a thread titled "Vacuum Distillation Column: Design And Control". The thread took about 37 posts to make some points regarding vacuum design as clear as I could. Air leakage was only one of them. If you research and read this thread, I think you will find common issues covered.

There are many other issues besides the air leakage. The selection of the vacuum equipment is sometimes a vital factor - or a vital mistake. The employment of Liquid Ring Pumps (LRPs) is highly desirable in some instances - and highly UNdesirable in others. I have had to troubleshoot LRP installations that simply were not suitable for that type of equipment. The application engineer simply forgot that the the principle of the LRP depends primarily on the ability to maintain a LIQUID piston -- and since that piston is a LIQUID, then it is subject to remaining as a liquid depending on its vapor pressure and the indicated temperature. The original design and patent for the LRP included mercury as the operating fluid medium. Today, mercury is not allowed in the vast majority of applications. Therefore, substitute fluids are employed - oils among some of them, and process fluids among others. Each one of these fluids must be capable of withstanding the vacuum condition generated without vaporizing excessively - i.e., having a very low vapor pressure. Gas binding (where the liquid ring starts to vaporize) has been prevalent in some applications.

Air infiltration is something that I have always fought - both at the design and operating level. My immediate resolution when dealing with low vacuums is to eradicate as many flanged connections as I possible can justify. Screwed fitting are not permitted. 100% welding is to be sought whereever possible. I can go on and on with the importance of the mechanical design and why it makes a difference. As DRS states, one never under-designs the capacity of a vacuum system. This is asking for trouble. A vacuum maintenance system should inherently be OVERSIZED - how much, is a matter of professional engineering judgment. This is the only practical way to design such a process.

Perhaps if you furnish us with specifics and all the basic data, we can delve into each one of your concerns in detail, but we need the specifics to go into that.

[SPOOK]

Thanks a lot ART for putting in so much effort to reply. really appreciated.

As far as the design calculations are concerned it has been repeatedly checked by my colleagues and myself so i am pretty sure about the accuracy of the math. What i meant by asking for your opinion was, for the system of 7 m3 and around 8 nozzles seven of them closed ranging between 2 inch to 6 inch in size, that the value of air leakage rate that I ended up with(15 kg/hr) did it sound practical enough.

I would not expect you to check my calculations and the methodology.

having such vast practical experience you can surely judge whether the leakage rate is reasonable enough and that is what i wanted to know.


As far as the details about the liquid ring vaccum pump is concerned, thank you for the insights into such system and the same is already under consideration for my system in particular. I will keep the details in mind while choosing a appropriate system for evacuation.

Also, the above discussions are compelling me to deduce that there is only one general design guidelines to calculate the air leakage for designing a vaccum system (correct me if i am wrong).
DRS also agrees with me regarding the above fact as he quoted,"In fact, it is the only resource I saw in about 14 years that deal with design of vacuum systems and absolutely free of cost"


Kindly give your insights.

REgards
Spook

[DRS]

Mr Montemayor,

That work book is another great contribution from you. When you provide fundamentals to support the calculation, this fairly reduces the chances of wrong calculation and I liked it the most. I hope, the OP will have a clear idea after going through the entire work book, word by word.

You mentioned about mercury being thought of as a sealing fluid. This was one aspect I thought during my initial days and didn't pursue further as it occured silly to me, somehow . May be now it is a purely acedamic exercise, still I would like to know more about it. Do you have any references for this?

[DRS]

Hi Spook,

Missed out your last post as we might have been writing simultaneously. I would like to caution you and others who read my post that I am basically a mechanical engineer and my knowledge is limited to my experience (2 years in Dye Intermediates, Sulphuric Acid Plant and Fertilizers and 5 years in API). Though I try to suggest something if only I know it properly, you can't take it for granted (unlike with other chemical wizards here) unless you review it with respect to the particular system under discussion.

First of all, I never came across vacuum systems in our cirriculum and I started learning about them after I started my job. It is during the time I shifted to Formulation/Biotech field, I came across E.Ludwig's books and still refer to them for some very specific issues and have deep respect about it like all other chemical engineers. That may be the reason why I was unaware of vacuum system design for considerable time of my experience.

Coming to your question, considering a volume of 7 cu.mtr, 15kg/hr may be a bit high but what Mr Montemayor sys (as I understood it) is that better be with redundancy when we actuall can't establish some data with respect to any specific system. Further, in the absence of all conditions pertaining to your system, one can't suggest the exact requirement that suits your problem.

There are certain occassions where I couldn't establish leak rate so I overdesigned the system and took care of it. Pharmaceutical autoclaves do require 1.3mBar/10 min as maximum possible leakage (which translates into about 0.012 cu.ft/min for a chamber volume of 96 cu.ft) and here we take care by tighter system and reducing the distance between autoclave and vacuum pump etc.

I see couple of issues with your system. You seem to be calculating the capacity of the pump by taking empty volume of the system and leakage rate into consideration. However, you seem to be ignoring vapor generation. If the water temperature is continuously being maintained at 90C, you have considerable vapor generation. If it is not maintained at 90C, the trick will be to maintain 20 mmHg (g) and yet remove the vapors from the space. Secondly, it is not mentioned as to what kind of control you are going to use to maintain the vessel at 20 mmHg (g). What if the pressure goes down further and will it cause trouble to your process. For LRPs you can open a bleed valve corresponding to the set vacuum and operate but how it is to be maintained for a steam jet vacuum system is unknown to me.

In any case, even with 15kg/h design leak rate, you will end up with a small vacuum pump (if it is LRP) and won't cost you much.

If you can provide your detailed calculation and the process details, it may help the members to specifically look into.

[SPOOK]

Thanks for your detailed response.

first of all my system is not very complicated and critical as well. but the main objective behind putting up such post was to find out what sort of leakage rates do process engineers end up with whey they have a system of a tank gross volume of 7 m3 with 8 nozzles on top of it seven of them close. as the requirement of vaccum in my system is not critical as it is required only to drive the vapors, generated inside the vessel due to high water temperature, out of the vessel.

you are absolutely write regarding the calculation of capacity of the pump. But the vapor generation has not been neglected when I specify the load of the vaccum system; I am specifying the non-condensable and condensable loads separately. What ever vapor is generated will form part of condensable load. And the air leakage will be the non-condensable load. as far as the control of the vaccum is concerned let me remind you that we are diverting away from the main cause of concern "the issue is reasonable values of leakage rates and calculation methods available"

i appreciate that a LRP is a very suitable option that is already under consideration. But I have ended up with an unreasonable size if I consider a water jet for the same application.



Regards
Spook