Well, not really on the cheap, not willing to cut corners that will cost me dearly down the road, but if I can cut a corner that won't be too expensive later on, what the heck.  (Warning, long post)  Anyway, a little about me.  First time home owner, lots of grand plans and such.  The house was built in 84, 2k sq ft, 4 br, 1 acre, county, and i have a creek in the back yard.  The creek is roughly 30 feet wide and the bank is 8 feet tall, but the water is usually only 3 feet deep.  During flood conditions, it usually spills over into the yard a little, no biggie.  Typical measured summer temp is 52F and winter is 45F.

Current HVAC system is a typical 3.5 ton 13 SEER Goodman heat pump split system, single return upstairs (12x25, way too small) and it handles the heat load just fine.  But knowing me, I cannot leave well enough alone.

Agreement with my wife:  The current system stays untouched and I can do what ever I want with geothermal.  At the thermostat location, there will be a switch to enable the desired system, split AtA or GT.

Please note that I am NOT doing a multistage setup with the 2 systems.  Its one or the other.

So here is my plan.  I am asking you guys for some advise, criticism, sanity checks, laughs, flames, rolling of eyes, whatever you deem necessary.

I was given a 3 ton Addison water source heat pump (previously used for GT also but well, pond discharge) and plan on basing everything around it.

I plan on piping both the suction and discharges of both units (split and GT) together but there will be dampers to prevent back flow (the T-stat and selector will also interlock out the unused system)

In the air return line before both units, I plan on putting a 7 ton evaporator trane vertical flat coil that I have already modified for chill water service. (removed TXV and small tubes, replaced with large bore).  According to calculations, should be able to get about 1 .5 ton of cooling with 50F water but minimal humidification. 

Install large filter box before flat coil that will hold 2ea 20x20 pleated filters.  Remove filter from existing housing but it will still be the system return.  The point here is to get lower filter face velocity.

Run 3ea 1.5in PVC pipes to the creek.  The reason for the large bore is due to the distance, 280 feet, planning for 300, which will only cost me ~9.5ft of head pressure loss at 15 GPM.  (I know I really only need 3 GPM per ton but I am planning for future additions) (The extra pipe is a backup to get the system back online in a hurry)  Should these pipes be buried in the same trench?  Should I isolate them with a few inches of soil from each other?  What do you think of PVC over PE?  I am not a fan of PE, too expensive and not as easy to get.

I also plan on putting a pair of 1.5kw electric heat elements in the water line going into the heat pump.  The theory is that if the water is to cold, I fall into a low COP area so by adding a small amount of heat, I can boost the unit back into its more efficient range.

Staging.  When in regular split AtA mode, there is no staging.  When GT mode, I think I want to run it like this:  First stage simply runs the blower and circulates water in the flat coil.  If that does not satisfy the demand, then it will bypass the flat coil and start the compressor on the GT then shut down once satisfied.  In heating mode, its only a single stage but it will have a temp sensor on the water output line to turn on the booster heaters one at a time.

Creek:  This is where it is going to be fun.  Because the creek goes from calm to flooding often, I cannot anchor the typical 300ft of 3/4" PE pipe per ton to the creek bottom.  That and the environmental people would be all over me.  Plan is to make 3 cells, 20 feet long, side by side, constructed as follows.  Put 2ea 3/4" copper pipes inside a 4in PVC pipe and on each end, put a 6" to 4" bell reducer to increase velocity to help prevent sediment build up.  There would be appropriate screens installed.  The loop would run likely run a 20% propylene glycol mixture.  I am concerned with possible embrittlement caused by propylene glycol and PVC, thats why the low concentration.

So what do you guys think?  Ive already started to gather materials and parts but still have a way to go.

There is no point in electrically heating entering water so as to boost COP. Better to live with the low COP and use resistance heat to warm the supply air.

PVC might work but PE is preferred as it is available in very long lengths and when properly fused is less likely to leak than a PVC line with many many joints.

You may need some sort of control, such as a humidistat or thermidistat to bring on the conventional unit if humidity gets high even if dry bulb temperature is at setpoint

Also beware creating too much restriction to airflow. You might consider investing in a manometer and learning how to use it.

If you see this through you'll learn a world of knowledge on HVAC and psychrometrics. Unless you are both lucky and persistent, I don't guess you'll save much money.

I used to do stuff like that but I was single at the time so only answered to myself for the various floods, freezeups and fires that occasionally resulted. I admire your wife's frontier spirit...

Good luck, do keep us advised...

There are lots of regulations regarding creeks (and they vary from state to state). They might not permit you to place anything in the creek that could be perceived to impede the flow.

Posted By engineer on 20 Apr 2010 02:02 PM
There is no point in electrically heating entering water so as to boost COP.

Not to mention that each 1.5 kW heater would only raise the EWT by
about 1°F at 9 GPM flow. The effect on COP would be microscopic.

FWIW, two of my property lines are small creeks (much smaller than
the one you describe) -- and so, I wish you GOOD LUCK in building
a submerged structure that'll survive a few summer thunderstorms.

...they don't call 'em "gully washers" for nothin',

Looby

If you are excavating 300' to the creek, why not just go closed loop?
j

Posted By joe.ami on 21 Apr 2010 11:59 PM
If you are excavating 300' to the creek, why not just go closed loop?
j

And bury part of the closed loop under the creek bed, sounds like a win-win

There are lots of regulations regarding creeks

Ive already shown my design to the warden and its been approved. Things are pretty lax here so as long as I do not make major changes to the watershed.

Posted By joe.ami on 21 Apr 2010 11:59 PM
If you are excavating 300' to the creek, why not just go closed loop?
j


And bury part of the closed loop under the creek bed, sounds like a win-win

It already is closed loop. I actually considered burying the pipe in the mud of the creek bed, it would silt over and have decent protection. My main concern would be future maintenance and if the warden would allow that. Ill have to ask.

Engineer, thanks for the tip about the humidity. It could become an issue.

Yes your design was already closed loop. I guess my point is why reinvent the wheel? You don't have to go 300' from your house to harvest required BTU's.
J

Interesting discussion.

I guess that digging 300' to the creek might be cheaper than excavating to bury a couple thousand feet of PE near the house. Plus the 45 to 52 degree running water in the creek bed will give better COP in both summer and winter than a loop field.

Gotta agree that the PVC sounds a little too fragile. Plain old PE would be much more sensible, but would a Slim Jim make some sense here?
(http://www.awebgeo.com/AWEB85c0.html?file=SlimJim.asp)

Why not use a regular two stage thermostat with Y1 hooked up to whichever unit is more efficient, and Y2 hooked up to the less efficient unit as your supplemental/backup heat?

Sounds like a fun project, if thats what your looking for. You may squeak a couple of extra degrees EWT out of this set up. Maybe, maybe not. I would prefer to just dig a trench and opt for no maintenance.

Slim Jim won't apply here, need constant depth.
I'm thinking among the characteristics of an intermittant stream are storm surges with lots of debris capable of causing harm to anything exposed within.
I like loops out of harms way.
J

What is a good place to buy PE pipe? Oddly enough, of all the places that sell PVC pipe, including the industrial wholesalers, home depot has the best price that I can find. When It comes to PE, I cannot find anything bigger than 3/4 in. Im looking for at least 1-1/4 (1-1/2 preferable) to cut down on head loss.

Posted By evil_admin on 27 Apr 2010 11:19 AM
I cannot find anything bigger than 3/4 in. Im looking for
at least 1-1/4 (1-1/2 preferable) to cut down on head loss.

Are you sure 1.5" would be preferable? Yes, bigger pipes
produces less head loss -- but also higher cost and less
turbulence (i.e., lower Reynolds# and possible reduction
in heat exchange efficiency). 1.5" HDPE sounds awfully
large for a 3-ton system.

experience: that which you have immediately after you needed it

Posted By Looby on 27 Apr 2010 04:43 PM

Posted By evil_admin on 27 Apr 2010 11:19 AM
I cannot find anything bigger than 3/4 in. Im looking for
at least 1-1/4 (1-1/2 preferable) to cut down on head loss.

Are you sure 1.5" would be preferable? Yes, bigger pipes
produces less head loss -- but also higher cost and less
turbulence (i.e., lower Reynolds# and possible reduction
in heat exchange efficiency). 1.5" HDPE sounds awfully
large for a 3-ton system.

experience: that which you have immediately after you needed it

Fittings go up astronomically as well on less commonly used sizes.
I'm suprised you are having trouble getting 1-1/4 for headers, on line or near home keep looking. Next you will need fusion tools to rent.
J

My reasoning for such large pipe is to get the head loss down so I could use a much smaller pump. If you guys think that I can go 300' each way with smaller, ill do it.

My flow calculations are based on a fluid with an SG of 1, elevation change of 0, fluw of 12 GPM and 300 feet each way.
1" 52' TDH or 22.5 PSI loss
1.25" 21' TDH or 9 PSI loss
1.5" 9' TDH or 4 PSI loss

If I am off, please tell me and Ill learn something.

This does not include the loss in the heat exchanger or the unit itself.

Why (in detail) is PVC not used? Ive seen several professionally installed systems where they used PVC to link the system to the shore, then use unions to connect to the PE coils in the lake. Ive installed PVC potable water lines that run well over 100PSI with no problem and seen some that have been in service for almost 20 years. Not trying to start a debate, just curious.

Finally, where would you guys suggest buying 1000 ft of PE pipe in either 1, 1.25 or 1.5 in?

It would be interesting to take the total pressure losses and convert to savings in electricity and then compare this to the pipe costs.

Only things I don't like about PVC generally is that it cracks when abused and low thermal conductivity.

Posted By evil_admin on 28 Apr 2010 11:27 AM
If you guys think that I can go 300' each way with smaller, ill do it

Designs by "you guys" on TheInternets are worth no more than
what you paid for 'em.

Elevation change has no effect on a closed loop system.

Don't worry 'bout SG, but DO worry about antifreeze type.
Antifreeze increases viscosity and greatly affects TDH.

Try this: http://waterfurnace.ca/Engineer/Misc%20References/Pressure%20Drop%20Calculation%20Instructions.pdf

"A little" overdesign can be extremely expensive -- do you
really want 4 gpm/ton? Compared to 3 gpm/ton, that's a 65%
increase in TDH -- and more than double the pumping power,
(for any given pipe diameter).

The smallest SlimJims are 2' tall and require 2' of unfrozen water above them so the 3' average depth of your stream prevents their use.

300' of trench would generally alllow the placement of about 3 tons worth of compact HDPE pipe in a "compact slinky" configuration and is an option that should be considered.

If you haven't lived owned the house for a few years (i.e. first time home buyer) you may want to make sure you are always going to have enough water flowing in the stream. To this end, you can find 30 years of streamflow gauge data for much of the U.S. at http://waterdata.usgs.gov/nwis/rt

The seasonal temperature variation of your stream suggests that it is primarily an effluent stream which recharges the regional water table and its baseflow is very dependent upon recent rainfall, snow melt etc. An influent stream, which gains water from the groundwater table will often fluctuate in baseflow and temperature significantly less and would be a better candidate.

-Adam

You guys have given me a lot to think about. And I greatly appreciate it.

"A little" overdesign can be extremely expensive -- do you
really want 4 gpm/ton? Compared to 3 gpm/ton, that's a 65%
increase in TDH -- and more than double the pumping power,
(for any given pipe diameter).

I am looking to add either an out building or an addition, that's the reason for the extra ton of water.

Down2Earth, can you tell me more about slimjims? Also, about this creek, its basically the runoff for the city. Ive carefully looked at the USGS waterdata site long before starting down this path and I know several people that live on its banks. I am pretty familiar with it. Im close to the middle of both of these stations so as to give you an idea of the creek.


http://waterdata.usgs.gov/tn/nwis/uv/?site_no=03535400&PARAmeter_cd=00065,00060
http://waterdata.usgs.gov/tn/nwis/uv/?site_no=03535200&PARAmeter_cd=00065,00060

I wonder if the Slim Jim requirement of 2' feet of unfrozen water above assumes static water (lake or pond).

If a Slim Jim were immersed in water with any significant flow I doubt it would need any depth above itself to transfer design heat.

Whatever heat exchanger is installed would have to deal with floods, debris and other effects of moving water, but if the water is truly always moving the heat exchanger could be much much smaller than that typically required in a pond / lake.