21 replies to this topic

[WORC]

Dear all,

 

Firstly thank you for this great site with infinite knowledge. I hope to learn and share more on this website. 

 

I'm a ChemE student working on our design project. For the project, we are exploring a novel modification to
an existing Organic Rankine Cycle (ORC) technology. In short, this Rankine cycle uses a caisson (basically a well drilled a few meters deep) to generate head instead of a pump. I’ve included the draft PFD of our system.

 

PFD:   PFD.pdf   10.52KB   28 downloads

 

As in our system, the heat source (geo-water) is coming into evaporator at 50000 m³/d (140ºC and 1400kPa). Heat transfers to and vaporizes our working fluid (Propane) which eventually rises and spins our gas expander to generate work. The expander outlet will then go through a condenser which cools working fluid to saturated liquid. Then it goes through a surge tank to the caisson again. This cycle repeats to generate work from low temperature heat source.

The objective of our project is to maximize the efficiency of the modified ORC system and compare it to a pump dependent system.

 

The questions:

1.   
It was quite confusing where to start in this cycle as every point depends on the previous unit/state.Any suggestions on how/where to start our equipment sizing calculations will be appreciated.

 

2.   
Also how can we find optimal mass flow rate of the working fluid for our system? Do we assume some value for this? If yes, then what could be the appropriate value?

 

3.   
In this process, we introduced the surge tank mainly for maintenance (refill, flow control, etc.) and safety purpose. How can we determine the best type of tank for this purpose (eg. Floating roof tank, Fixed Roof Tanks,…). Also how can we determine the height to diameter ratio? (eg. Max for floating roof tank was given as 2:1 for height:dia)

 

4.   
Expander type: some ORC specific type expanders were looked at. For example scroll type and screw type etc. Our scope was not to focus on the mechanical detailing of these but solely the performance of these. Is there any specific advantages/disadvantages on these expander? And what should I be concentrating on when performing design calculations on gas expanders?

 

We value your advice and suggestions.

Thank you for guiding us.

 

 

Vig from (Team WORC)  

[thorium90]

Well, you could start with stream 5. Assume some starting value. You probably know enough about stream 6 and H101 to calculate stream 1. Continue your balances until you get back to stream 5. If it is different, use this new value and do all the calculations again to get a new value for stream 5. Keep doing until the previous and current value of stream 5 are very similar

 

Optimal wrt? How do you define your optimization objective function? What are the terms you put in it and will it be a minimization or maximization function? Eg: minimize capital and operating costs or maximize profit.

 

I assume there will still be some pressure remaining, so it will be some pressure vessel, and not an actual storage tank. (Propane has bp of -42C at atm pressure). You can make some assumptions for an LD ratio, some books and some threads on the forum has some guides

Edited by thorium90, 18 February 2013 - 11:31 PM.

[Art Montemayor]

WORC:
 

I don’t know what level you are in your Chem E. curriculum, but by the nature of your questions, you don’t seem to have a thorough grasp of what is going on within the Rankine cycle.  I strongly recommend you go back to basics and study the details of the basic Rankine cycle.  This is a very important issue and one you should completely understand for the sake of excelling in your future Chem E courses.  I am going to answer your questions in a manner meant to stimulate you in that direction, hoping you progress towards mastering this part of thermodynamics:

 

It was quite confusing where to start in this cycle as every point depends on the previous unit/state.  Any suggestions on how/where to start our equipment sizing calculations will be appreciated.

In the Rankine cycle you are basically carrying out two principle unit operations in a closed loop: (1) you are vaporizing a working fluid; and (2) you are condensing the same fluid.  In order to maintain your closed loop (a major priority), you must have the appropriate heat input that will vaporize the fluid and a condensing medium that will allow its subsequent condensation.  Therefore, by common logic and sense, you must supply the heat source that furnishes sufficient temperature to vaporize and the condensing medium temperature that allows condensation at the expansion turbine’s exhaust pressure.  (you could employ a condensing turbine, but that is another topic.)  therefore, the first things you must fix and/or identify are the vaporization and condensation temperatures in accordance with available vaporization and condensation resources.

 

Also how can we find optimal mass flow rate of the working fluid for our system?  Do we assume some value for this?  If yes, then what could be the appropriate value?

I do not believe there is an “optimization” – as such – involved in your application.  The working fluid’s flow rate is fixed by the availability of your heat source – which you have identified.  To the extent that you can successfully transfer heat from your geo-water to your working fluid, you will fix the amount of vaporized working fluid going to the expansion turbine.
 

In this process, we introduced the surge tank mainly for maintenance (refill, flow control, etc.) and safety purpose.  How can we determine the best type of tank for this purpose (eg. Floating roof tank, Fixed Roof Tanks,…).  Also how can we determine the height to diameter ratio? (eg. Max for floating roof tank was given as 2:1 for height:dia)

Why do ChE students get so “hung up” on a pressure vessel’s height-to-diameter ratio?  This is a mundane value and will be fixed by your requirements per the location or by the vessel fabricator in accordance with your data sheet.  Why are you using a floating roof storage tank?  Determine the capacity you need as per your required inventory, residence time, instrumentation requirements, and common sense.  The best dimensions are those that comply with your installation and operating requirements together with the availability of standard-sized steel plates as used by the fabricator.
 

Expander type: some ORC specific type expanders were looked at.  For example scroll type and screw type etc.  Our scope was not to focus on the mechanical detailing of these but solely the performance of these.  Is there any specific advantages/disadvantages on these expander?  And what should I be concentrating on when performing design calculations on gas expanders?

Why are you debating on using a scroll or screw type of expander?  Aren’t these normally compression  devices instead of expansion?  Why not apply a basic centrifugal-turboexpander type?  That’s what I would look at first.

 

Look at the attached PowerPoint file and see how you can conceptually improve the efficiency of a Rankine Cycle.  A Rankine Cycle is the same regardless of the working fluid employed – “organic” or otherwise.  The organic version of the Rankine Cycle enters the picture only for marketing purposes and makes no difference on the thermodynamics of the application.

 

How do you conceive that you can substitute a hydrostatic liquid head for a conventional pump within the Rankine Cycle?  Can you explain how a “a few meters” of depth can generate head instead of a pump?

 

What are you proposing to use as your cool medium to carry out your condensing duty?

 

[thorium90]

As Art has pointed out, it seems impossible your evaporator can possibly heat the subcooled propane from the surge tank to saturation and to superheat all together,

Assuming you had more than one exchanger there, after the expander, the reduced pressure would require cooling the propane pretty cold to condense it which will require chilling fluid. (assuming you choose a condensing turbine)

It would appear that the version with the pump is the only workable version

 

Perhaps WORC could define the expected stream temperatures and pressures in the PFD? Would help in understanding its feasibility

 

A quick search on google does show that screw and scroll type expanders do exist and work the opposite way from their compressor counterparts.

I think Art missed out the powerpoint file he mentioned.

Edited by thorium90, 19 February 2013 - 01:26 PM.

[WORC]

Hi Art and thorium90,

Thanks for your insight into the topic. This is actually our final year project. Sorry I should have been clearer, as for the question “where to start” was actually with the caisson limitations. Since the heat exchanger is going to be down at the bottom of the caisson, it has its limitations with the diameter. And ultimately it affects the entire cycle. The cost of drilling a caisson grows exponentially with diameter. However, we still haven’t got the pricing data yet. Therefore we were initially thinking about having 24” dia.

In addition, almost every article we came across were using scroll expanders for ORC systems and some suggests positive displacement expanders(scroll and screw type) give better efficiency than turbo expanders as these ORC systems are small scale units. This was the main reason why we were looking at scroll and screw expanders. We probably got driven away by the literature search for the expander type. But I’ll start looking into centrifugal-turboexpander as per your suggestion.

Since the work generated is driven by pressure difference and as we are using propane as our working fluid, we decided to start off with 500-700m deep for the caisson. Which generates pressure of 24 bar to 34 bar (calculated using density*g *h). Propane have a very low boiling point at atm pressure (-42.09ºC) so we need high pressure here so that our condenser (described below) could handle the load and remove some sensible heat and latent heat to perform phase change to liquid propane. This liquid propane flows back into caisson and cycles.

Caisson:

-behaves more like a well
 

24” dia

500-700m deep

Contains heat exchanger (column shell and tube) and piping (heat source and working fluid)

Condenser:

We are evaluating three systems:
1. Water cooling tower (@ 20ºC)
2. Aerial cooler
3. Hybrid (both water and aerial cooler)

 

Why the hydrostatic liquid head can work: We believe there is a pressure difference between the liquid column (flowing down the caisson) and the gas column (flowing up the evaporator) which could drive the cycle.

We are already doing energy calculations at the evaporator and condenser to determine some of the variables. As for our optimization function, it would be to maximize the (work generated/capital cost).

The powerpoint doc was not attached. It will be great if you could post that again. Thanks again.

 

Regards,

WORC

[thorium90]

A quick calculation shows some possiblity as the pressure drop of gas back up the caisson still gives enough pressure remaining. About 20-30C superheat seems possible. However it is noted that to use water or air cooling, expander outlet would have to be >10bar.

Edited by thorium90, 19 February 2013 - 02:37 PM.

[WORC]

Hi again,

probably just missed thorium90's message before my previous post. But I’ve attached the details of the streams. But all the values in this stream are just preliminary and quick calculation. We would have to do further analysis on these, but I believe it serves the purpose to understand the project. But one problem arise is the heat exchanger (evaporator) at the caisson's bottom was found to have 3400m2 with 90 tubes approximately. This is bad as we need a really really long series of heat exchangers in the small diameter caisson.

 

 PFD_updated with state conditions_v1.png   38.43KB   17 downloads (Corrected version)

 

Thorium90: yes, you are right, we are intending to have high pressure(~11bar) out the expander to cool this using cooling water and air cooler. Is the pressure drop in the vapor from the evaporator to the expander significant?

Also to counter the series heat exchangers problem, we were thinking to have a preheater along the pipe flowing down the caisson. I'm guessing we might have to tweek our mass flowrate to adjust for this reason, as all we need is to vaporize the propane and superheat a bit. Any suggestions would be great.

 

Thanks,

WORC

Edited by WORC, 21 February 2013 - 01:07 AM.

[thorium90]

Its odd that stream 2 after the expander has the same temperature as stream 1. Your energy balance calculations are quite off. It would appear large pipes 24" are needed to mitigate the pressure drop. Alot of space would be required to build all that exchangers 700m underground. There would be quite alot of pressure drop back up the pipe and you still need to have ~11bar after the expander. The cost of this project would seem uneconomical as there doesnt seem to be alot of power left to turn the turbines.

Edited by thorium90, 20 February 2013 - 03:47 AM.

[WORC]

Hi thorium90,

 

It's a mistake there with stream 1 and 2. It should be less (~35 to 45 deg C).

Opps, I thought the density of propane vapor is going to be low but it is indeed quite high (at P=30bar, T=90deg C) it is coming at 66.4kg/m3. That means I'll get a pressure drop of 4.55 bar (15% loss). Making the stream to the expander to be at ~26bar. And adding the friction factor will lower it even more. If I drill deeper than it gets more dense (if I don't then less pressure gain) and my heat source is not high enough to superheat it to high T. I'm stuck. Any work around?

 

If I assume 26 bar inlet to the expander, then the expansion ratio have to be ~ 2.4 in order to get the outlet at 11bar. I'll have to check if this is enough to generate a decent work.

 

As for the heat exchangers, I think we can drill it with a bigger hole (maybe even 2m dia at max) but that would be expensive and it also mean the caisson can't go as deep as 700 m. 

 

As we are bringing the propane near critical conditions  (Tc=96.65 C, Pc=4.24MPa), the vaporization energy(just adding the latent heat) will be lower compare to raising the temperature from ~30C to 76C. So is it worth to have a preheater along the pipe to the bottom of the caisson, so that it reduces the load on the evaporator?

 

Thanks,

WORC

[thorium90]

Some kind of preheat is possible before the evaporator but you cant boil it off before it reaches the bottom. I get about 4+bar for the pressure drop back up too. Your water is at 140C, I think its still possible for more superheat but the available pressure appears to present a problem. I think your electricity costs from this plant will turn out to be quite high...

[WORC]

Hi thorium90,

 

I certainly agree with you. The price of electricity costs is going to be high (current electricity price is low). But one good thing about this system is it gets work and energy from environment (eliminates pump) and everything else is capital cost. Which also means the payback would be much more longer.

 

I have a question for the preheater:

By simulation it was found that preheater saves a whole lot of evaporator area.

I was intending to use the reduced heat source (out of the evaporator) in a double pipe exchanger. But most of the double pipe exchanger I came across are in 'U' forms as shown in the picture below. I guess these are designed in U forms so as to save space. So can I actually have this HE designed in a straight pipe fashion (pipe inside a bigger pipe) without loops or the "U" for this purpose. I calculated the area required for this counter current preheater to be ~0.1202 m2 and by have a 20cm dia tube, it came ~382m long. I believe It cannot be fabricated as a single piece, but can smaller double pipe exchangers be flange together for this purpose? Please advice. Thank you.

 

Preheater Details:

using U=650

Th1 - hot side inlet being 140 deg C

Tc1 - cold propane @ 27deg C

fixing Tc2 - preheated propane @ 45deg C

 

 

 Double-Pipe_Heat_Exchanger.png   7.46KB   2 downloads

 

 

[thorium90]

The reason you get such a long tube is because double pipe exchangers are not suitable for these applications. They are suitable for low flow rate or high pressures. They are for easy installation and maintenance, not really for boiling of propane..... Perhaps you can consider the more classic shell and tube type?

There would have to be some kind of control to ensure that it does not boil off on the way down. It might be more suitable for the preheater to be just above the main heater at the bottom. And the heater would use the remaining heat energy from the water after the main heater.

Edited by thorium90, 21 February 2013 - 04:03 AM.

[WORC]

hi thorium90,

 

Now we got the typical size and cost data for our caisson. It is giving new challenges, the caisson existing today can only go ~150m deep (maybe we can dig in 50m more) with around 0.4m dia. For larger dia like around 3m it can go shallow with a max of around 40m deep.

 

In other words, this does not provide enough head (initially we assume 500m deep giving around ~24bar and with initial pressure (~33bar in total)). Now with 200m deep  a total of only 20bar can be accomplished. Which means our expander work output may not be feasible as we still need to retain 11 bars to liquefy propane using water or aerial cooler.

 

Another problem with the shell and tube heat exchanger, the phase separator demands for a larger diameter (interfacial area between vapor and liquid). It like around ~2-3m for a flow rate of 70-80kg propane/s. This flow rate is the minimum that gives some decent work output. 

 

Any suggestions as how to raise the hydro-static pressure. Propane is specially preferred for this study. One way I thought is to increase the pressure is by mixing propane with another solution with higher density. But I could not imagine how that would work. Any thoughts?

 

Thanks for your help.

 

WORC

[thorium90]

At 200m, i estimate a pressure of only about 16-17bar back up at the surface. Conservatively speaking, only about 1MW of power production. Phase separator? 2-3m in height? What do you mean you need a phase separator?

[WORC]

Sorry, by phase separator I meant the evaporator at the bottom. Firstly using the Souders-Brown equation, the allowable gas velocity was found. Then the cross sectional area of the evaporator was found using:

 

A=m/(gas density * allowable gas velocity)

 

and the dia was found to be 2-3m. So is there any way the pressure could be increased for a 200m caisson? Does mixing propane with another liquid help or make thing worse?

 

WORC

[thorium90]

The Souders-Brown equation is not used in the calculation of a shell and tube exchanger with water in the tube and propane in the shell side with baffles and all.....

Im not sure what you can mix with that can help, you can read around... Otherwise, you will need a pump, but that would make the economics worse..

[WORC]

Why isn't the Souders-Brown equation used in Shell-tube exchanger? Isn't the heat exchanger acting as evaporator? Is there an alternative equation to determine the interfacial area for Shell-tube exchanger?

Thanks.

 

WORC

[thorium90]

This might be a good read.

Attached Files

•  Effectively design shell n tube exchanger.pdf   241.91KB   11 downloads

[WORC]

Hi,

 

As for the pressure vessel (pls refer to #7 PFD), it was intended to act as a storage system during maintenance and also assist in the flow control down the caisson. However in order for it to act as a storage system, it needs to be filled partially with "something" during operation so that when it's time for maintenance, that "something" can be taken out and this working fluid can fill the vessel. The problem now is what should the "something" be. One idea was to fill it with a gas and purge it later when we want to fill the vessel. Mainly gas as we don't want this "something" to flow down the caisson with our working fluid. Any suggestions?

 

Note: pressure vessel was chosen over storage tanks especially because there is a huge pressure left in the stream and we don't want to lose it.

 

Thank you.

 

WORC

ps: I hope I'm clear, if not pls let me know I will try to explain it better.

[thorium90]

I understand it is a closed system. During operation it will hold propane. During maintenance of the vessel,, u probably need to drain it out to a secondary vessel so you can work on it. If the maintenance does not involve opening it then you can keep it isolated. Some amount of refrigeration might be needed.

 

Btw, you do realise that since it is a closed system, that "needs to be filled partially with "something" during operation" means it is your working fluid, aka: propane inside during operation?

Edited by thorium90, 14 March 2013 - 10:53 AM.

[WORC]

Yes, but initially I was thinking about having a bladder or piston accumulator with some compressible gas. But the liq-vapor equilibrium of propane in the tank should be able to keep the vessel pressurized, right? Maybe I need a bypass from the line before condenser (propane vapor state) to the vessel to maintain the pressure in the vessel.

 

WORC

[thorium90]

The purpose of a bladder or piston accumulator is not to store liquid propane during shutdown maintenance. Either I've understood wrong or you are really getting off topic for your design.