Skips new Seafrost poor performance problem

[Richard Kollmann]

Skip, I am trying to follow your refrigeration troubles but lack the normal information needed to formulate an opinion that might help you. Information on Seafrost 12/24 volt units I have heard from boaters has been good although this is only three or four Emails. Your first system was a Frigoboat keel cooler with a BD50 compressor that performed well at freezer and refrigerator temperatures until system failed. You now report new SeaFrost air and water cooled unit with a 30% larger BD80 compressor will not maintain previous Frigoboat BD50 compressor box temperatures with either air or water cooling.

As I understand it your new BD80 system is powering two or three evaporator thermo plates not holding plates or roll-bond thin plate evaporators.. Refrigerant flow control is accomplishes by one or more evaporator pressure regulators instead of a capillary tube or TXV.


The poorer performance change from the new BD80 compared to the old BD50 Must result from one of the following:

• Insulation has lost some of its original R value. After 10 years this is possible on open cell insulation without moisture shielding but not a problem on your Extruded Polystyrene board. Morgan OI boas did have a problem with moisture interring insulation from a leaking deck seal this question can be eliminated by drilling a small hole inside box near bottom of liner. If dry behind liner moisture is not a performance problem.
• Under the summer climate conditions where the boat is now ambient water and air temperatures are about the same so why is it this unit with an oversize fan condenser is not performing as well or better with air cooling? Normally boaters find keel coolers are less energy efficient in seawater above 85 degrees F over a 24 hour day than air fan cooled?
• It is possible that refrigerant is contaminated. I found one boater that spent over $1,000 on service calls because another manufacturer failed to remove test nitrogen from either the new condensing unit or evaporator assembly.
• Refrigerant flow control is always a possible destroyer of good refrigeration performance. Of all the different refrigerant flow controls I find the suction pressure regulator devise the most difficult to be precise with when balancing a boat refrigeration system. These small pressure regulation devices were designed for drink machines where ice was produced in one refrigerant circuit and cooling liquid in another. When pressure regulator are used at freezer flow temperatures valve adjustments are even more delicate. Capillary tubes have a narrow flow control range, Thermo Expansion Valves used in mobile boat refrigeration have a much wider flow control range while Suction Pressure regulators are manually adjusted to produce a fixed flow control range like 4 psi to 10 psi.

If you could find a qualified technician and not one from HELL he will be able to analyze and see if high and low pressures are in line with 134a refrigerants for desired evaporator temperatures. Unfortunately these small pressure regulators do not have a low pressure test port as do the larger valves, sometimes used on larger holding plate systems. There are no two ice box conversion system installed in exactly the same way so tubing lengths, number of bends will all change the actual evaporator super heat by reading low pressure at compressor service port instead of valve complicating fixed pressure valve adjustments.

Skip, I do not have an answer to the question about how to handle seawater cooling system flow restrictions.

Hopefully this new thread will get one or two responders with SeaFrost new BD80 compressor systems to respond.

[ka4wja]

Richard,

I just wanted to say Thank You!
I have learned a lot from reading your books and some of your posts on refrigeration....
Whether you end up getting Skip's issues straightened out or not, I just wanted to say thanks, from so many of us who have appreciated all your efforts!
(in my opinion, you are "the man" of marine refrigeration!)

Fair winds!

John
s/v Annie Laurie

[Vasco]

Originally Posted by ka4wja

Richard,

I just wanted to say Thank You!
I have learned a lot from reading your books and some of your posts on refrigeration....
Whether you end up getting Skip's issues straightened out or not, I just wanted to say thanks, from so many of us who have appreciated all your efforts!
(in my opinion, you are "the man" of marine refrigeration!)

[Celestialsailor]

Yup...same here. You are a huge asset to this forum.

[skipgundlach]

Hello, Richard,

Originally Posted by Richard Kollmann

Skip, I am
trying to follow your refrigeration troubles but lack the normal information
needed to formulate an opinion that might help you. Information on Seafrost
12/24 volt units I have heard from boaters has been good although this is
only three or four Emails. Your first system was a Frigoboat keel cooler
with a BD50 compressor that performed well at freezer and refrigerator
temperatures until system failed. You now report new SeaFrost air and water
cooled unit with a 30% larger BD80 compressor will not maintain previous
Frigoboat BD50 compressor box temperatures with either air or water
cooling.

Almost right - there is no way to do water-only cooling. However, even on
high, non-stop (I looked up some prior correspondence with Cleave - Sea
Frost, and Clay - Hansen Marine), air cooling only will not succeed in
cycling the freezer, and temperatures gradually rise to barely frozen; as a
result, also, my reefer rises and can't keep up with the near-freezing input
vs the usual well-under-freezing input..

Originally Posted by Richard Kollmann

dd


As I understand it your new BD80 system is
powering two or three evaporator thermo plates not holding plates or
roll-bond thin plate evaporators.. Refrigerant flow control is accomplishes
by one or more evaporator pressure regulators instead of a capillary tube or
TXV.

Yes. My system is like as shown in their website, with, in our case,
instead of the two larger plates which, in the end, didn't fit due to the
slight tapering between top and bottom, originally ordered, Clay went with 3
slightly smaller plates, which are mounted on the walls not of the spillover
divider. Without doing the math, I expect that this coverage meets the
specified minimum of 60% coverage of the available wall surfaces, as the
plates could have been longer if needed. As it is, they extend well under
the shelf position we have.

BD Series

We have the largest model, the BDXPXAW System. As you can see in their
brochure, and the link to the condenser test (which presumes to demonstrate
the superiority of air over keel cooling, and describes their water cooling
system, which appears to be merely a single loop), we have an oversized air
condenser, and a water jacket condenser. I confess to not having
taken detailed pictures of the unit when it was open during installation, so
I don't really remember the layout in it. During installation, it was
presumed that we would be operating on air cooling only, but Clay ordered
the water option unit which now has a rocker rather than the shown toggle
switch for enabling the pump, presumed changed for durability reasons.

The main page also has amp draws for each of their units, as well as the
available BTUs removed. Note that it's either at -10F or -30F plate
temperatures. Given that in my conversations with Cleave, wherein he
suggested using whatever would produce something in the +20F range (which
would mean plate temperatures of about a 15F shutoff), I think that's a bit
disingenuous. The BEST, with water and air, that I could not-quite reach is
a plate temp of +2.5.

The anectdotal claims of too many to count here in CF of box temperatures of
zero or lower make me think it must be possible. Just not with my
system/installation/environment, because we've been having this discussion
since we left St. Augustine, with the above results.

Our system has an AEO which - if you can believe the literature, is very
smart, adjusting RPM by 400 per time, up or down as needed to achieve a 52%
run time. Regardless of setting (low, medium, high, automatic), I've never
experienced a 52% runtime. As you know, for a while I was being entirely
anal - well over a thousand readings - about recording times and
temperatures of on-off moments (one reading when it went off, one when it
restarted) for both the refrigerator and freezer; about the best I achieved
was a 2/3, and more like 80% runtime vs offtime. Our control is the ETT.

In conversation during Clay's fiddling with the system, he told me that
Cleave recommended not using the ETT in automatic, and that he just left HIS
on low all the time. Eventually, recently, I got the system to - at least -
cycle reliably (with water cooling working) on medium. In conversation with
Cleave, about using automatic, which was presumed to be the most efficient,
he suggested running it on medium, as that setting used a bit less power
than automatic. The inference was that this was overall - that using
automatic cost more than straight medium, which, by extrapolation, was way
more than straight low (which can't keep up running full time). As this was
all by phone, I have no written confirmation of that, but it appears moot,
as low only won't cut it.

I haven't, in a great while, done overnight calculations of draw. I guess
it's time to do that again. What I do is note our current AH deficiency when
we turn off the light (two 0.5A fans and miscellaneous overhead the only
load, dark and still so no wind or solar contribution), and again when we
get up, before turning on any other loads. For weeks on end, the AH
consumed were consistent when regarded as an average over the hours
measured.

From that I infer - without having a dedicated amphour recorder attached to
the unit/breaker - that our system behaves consistently and that I can rely
on those figures.

Our system has something not shown in the brochure - an
accumulator/filter/dryer with a sight glass. When the system was first
charged, there were bubbles; Clay said that was of no event. However,
eventually, during a roadtrip where we were gone for a month, he continued
to attempt to get performance from the system. The constant pressure valve,
on which more later, along with both temperature probes, were replaced. We
even tried using box temperature as the trigger, rather than plate
temperature, for a bit, during the early days, when I was still aboard.
While we were gone, apparently Clay capitulated, and installed the water
cooling pump and lines, eventually pronouncing it good, and increased the
charge to where there were only the most infrequent and minor bubbles in the
glass.

Finally, we have a constant pressure valve, nominally set (see below) for
4PSI, with which Clay fiddled fairly extensively before replacing the first
one, considering that to be defective.

I don't - and guess that Clay didn't, either - know whether our probes which
were the first defective replacements were, indeed, defective. Perhaps we
simply were not getting to temperatures. Both Clay and I have IR guns; they
don't agree with each other, or with indicated box temps. Out of
frustration, Clay bought a simple mechanical thermometer, and mounted it in
various places around the box during testing. If THAT could be believed,
the actual temperatures AT THE PROBE (hung it on the wire right before the
probe) were between 8-10 degrees colder than indicated on the ETT. Shooting
the probe temp supported that distinction.

So, I'm left with slowly raising the cutoff (now at +6.5 and a hystersis
of - also - 6.5), to achieve a box temperature which is nominally warmer
(nominally as we don't have the previous Carel thermostat and probe
installed for reference) than what the Frigoboat system did at half the
amps. During Clay's test, he bought some ice cream; until he fiddled some
more, it was not firm - which means, to me, that what I see for temperatures
on the ETT cannot be relied upon, and I'm constrained to dead reckoning to
attempt a box temperature which I used to be able - warm loading aside - to
keep within 2 degrees (8 cutoff, 2 hysterisis) in my prior system.

Granted, that was box, and not plate temperature, as I understand to be the
industry norm, so I have no idea what my actual plate temps were. However,
as it was of the thinwall (with a SS front for longevity) variety, I
speculate it changed temps rather more swiftly than the current plates,
which act a bit like cold plates; there's something inside the box, it
seems, which acts a bit as a thermal flywheel. If so, the plate temps, when
not charging, likely were very similar to box temps (of course, I could be
entirely mistaken, but logic suggests), and the ability to quickly remove
heat (because of the speed of decline in temps during evaporation) seems
like it would be more effective at maintaining that narrow hysterisis.

By contrast, my ETT box probe readings run about 13 or so cutoff, and low
20s restart, a huge swing by comparison. In one of the other threads,
someone suggested going to a 2-3° hysteresis for improvement - do you agree?

Finally, this increase in my cutoff temps has meant that my reefer actually
cycles - in rough approximation with the freezer, because when the spillover
fan engages, it blows the warmer air into the freezer, causing it to
energize soon after the reefer has shut off, usually - rather than was the
case when we were a couple of degrees lower. In that lower freezer temps,
the reefer was able to get by with convection, running the spillover fan
only when warm loaded, and, occasionally, the temp would drop below our
desired level.

So, from that (our temps are such that we actually have to use the spillover
fan on a regular basis), I infer that our system is reasonably balanced in
temperature terms. In addition, both boxes cycle reliably (the freezer
actually - if the water is working properly - gets to turn off on a regular
basis), albeit at a much higher energy cost than I'm happy about.

Originally Posted by Richard Kollmann

The poorer performance change from the new
BD80 compared to the old BD50 Must result from one of the
following:

• Insulation has lost some of its original R
  value. After 10 years this is possible on open cell insulation without
  moisture shielding but not a problem on your Extruded Polystyrene board.
  Morgan OI boas did have a problem with moisture interring insulation from a
  leaking deck seal this question can be eliminated by drilling a small hole
  inside box near bottom of liner. If dry behind liner moisture is not a
  performance problem.
  

• When the first plates were found to be too large to fit both into the box,
  we had a hole at the bottom of the first one mounted (and three others, too,
  of course) which was open. No water. We epoxy/cabosil filled those holes
  and started over for the new plates ordered. FWIW, Clay was able to sell
  those larger plates for an installation before we returned from our road
  trip, so he was particularly pleased at not having to have dead inventory...
  
  

  Originally Posted by Richard Kollmann

  Under the summer climate conditions where
  the boat is now ambient water and air temperatures are about the same so why
  is it this unit with an oversize fan condenser is not performing as well or
  better with air cooling? Normally boaters find keel coolers are less energy
  efficient in seawater above 85 degrees F over a 24 hour day than air fan
  cooled?
  

  That is, to put it mildly, the question of the day.
  
  

  Originally Posted by Richard Kollmann

  It is possible that refrigerant is
  contaminated. I found one boater that spent over $1,000 on service calls
  because another manufacturer failed to remove test nitrogen from either the
  new condensing unit or evaporator assembly.
  

  I'm not in a position to judge - but I'm confident that Clay knows what he
  was doing, and, this being a new system, I don't know that nitrogen was
  involved in it, ever. Additionally, I did see him do the initial evacuation
  test (leaving the gauges connected overnight), and since he replaced the
  constant pressure valve, I assume he did so again at that time. So, I am
  doubtful of that issue being the cause.
  
  

  Originally Posted by Richard Kollmann

  Refrigerant flow control is always a
  possible destroyer of good refrigeration performance. Of all the different
  refrigerant flow controls I find the suction pressure regulator devise the
  most difficult to be precise with when balancing a boat refrigeration
  system. These small pressure regulation devices were designed for drink
  machines where ice was produced in one refrigerant circuit and cooling
  liquid in another. When pressure regulator are used at freezer flow
  temperatures valve adjustments are even more delicate. Capillary tubes have
  a narrow flow control range, Thermo Expansion Valves used in mobile boat
  refrigeration have a much wider flow control range while Suction Pressure
  regulators are manually adjusted to produce a fixed flow control range like
  4 psi to 10 psi.

Ours is the suction pressure regulator variety. At one point, after a
decommissioning (we have to decommission the system every time we're off the
boat, as it would flatten our batteries in a few days without our being able
to recharge with the Honda 2000), I noted that the bottom half of the third
plate wasn't frosted. Conversation with Cleave had me fiddling with this
regulator.

His initial instruction was to mark the start point and give it a half turn
CW. That very quickly not only frosted the plate, but the line all the way
to the compressor. Over a series of steps, I reduced that change by half
each time. When I got to where it HAD been, it was still over-frosted. So, I
gave it a quarter turn CCW, and the frost receded. Too far, I did the
same - restoring by halves - and wound up where I started, which was,
however, by this time, proper - the plate was fully frosted, and the line,
as well as the immediate area (less than 1/4") of the plate from which it
exited, wasn't.

However, I've since defrosted; the restart has the bottom third of the third
plate with no frost, again. So, I've again begun having to fiddle with that
valve. I am not at all pleased to think that I will have to do this every
time I defrost or decommission (next decommission coming in a couple of
months).

So, I confirm your opinion of delicacy. However, I fail to understand why,
once set, the regulator should require messing with again, let alone every
time.

Originally Posted by Richard Kollmann

If you could find a qualified technician and
not one from HELL he will be able to analyze and see if high and low
pressures are in line with 134a refrigerants for desired evaporator
temperatures. Unfortunately these small pressure regulators do not have a
low pressure test port as do the larger valves, sometimes used on larger
holding plate systems. There are no two ice box conversion system installed
in exactly the same way so tubing lengths, number of bends will all change
the actual evaporator super heat by reading low pressure at compressor
service port instead of valve complicating fixed pressure valve
adjustments.

Would that be you? I will, eventually, get the rest of my game-stoppers
sorted out, and resume my shakedown; Lake Sylvia likely would be one of my
stops, as it was nearly 20 months ago when we started our shakedown from
hell. (Most of you don't know that EVERY time we moved the boat, there was
something which HAD to be resolved before we could continue. There have been
seven such events since we launched, and more have arrived since we tied up
again here in Vero Beach last Christmas...)

However, I can't get over the angui$h of losing a system which worked
perfectly until it didn't at all, and having $4K tied up in a system which
displays symptoms of not working properly. As seen in a different thread,
Cleave calculates that I have a BTU load well within the capability of our
system - but yet, suggests I'd have better results in total amp draw to
install yet another system for the reefer alone. Aside from the fact that I
can't afford it, I continue to be in the dark as to why a nominally higher
capacity system, using the vaunted, superior to water, air cooling, can't do
the job that a smaller compressor and inefficient keel cooler shrugged off.

Originally Posted by Richard Kollmann

Skip, I do not have an answer to the question
about how to handle seawater cooling system flow restrictions.


Hopefully this new thread will get one or two responders with SeaFrost new
BD80 compressor systems to respond.

Thanks for your help, Richard, as always. Sorry we missed each other last
time around.

L8R

Skip

[Richard Kollmann]

Skips Design

Skip, when you are looking for answers to a one of a kind refrigeration system problem and this is what you have even the manufacturer may not be able to help. And NO, I do not have a simple answer. You must forget everything you think you know and all advice good or bad and consider your system components from a functional application stand point because that is where the problem is.


Selecting Compressor Btu Capacity
You have a 10 cubic ft box divided to produce produced a freezer and refrigerator section. When previously spending a good deal of time in warm climates you established a base line of performance with a BD50 powered refrigeration system. After replacing old system with a 30% larger BD80 compressor system your question is why performance is not equal or better than the smaller capacity old failed unit.

Compressor Selection
Very few technicians understand the application and of a one of a kind installation I will try to break it down for you in hope it will allow you to improve present performance. Mobile refrigeration design begins with size of icebox to be converted, amount of insulation, temperatures desired to be maintained in box, and worst case climate condition where boat might possibly be cruising. With this information daily box heat load can be determined. But in your case you know from the base line you remember that BD50 was marginally satisfing the Btu capacity and desired box temperature in worst case ambient air and water conditions. The BD80 additional 300 Btu at plus 10degrees F over the BD50 provides a perfect choice based on old systems base line performance.

Installation of AEO Module
Adding the automatic compressor speed controller to BD80 compressor in theory will without it knowing what temperatures are anywhere throughout system may or may not optimize daily energy consumption. The thermostat always controls box temperature and indirectly controls compressor speed this can result in less desirable performance if other components in system do not have the ability to handle Btu output of compressor.

Condenser operation.
This BD compressor’s heated high pressure gas vapor must pass through a condenser with ability to lower super heated vapor not only enough to condense it to liquid but be able to handle maximum capacity flow of Btus from compressor at its highest Rpm. From a visual check of condenser cooling fins it looks sufficient enough to handle max capacity of a BD80 compressor in 90 degree ambient air conditions without loss of efficiency. You can always verify good condenser performance by high pressure reading of from 105 to 120 psi.

Evaporator Capacity
Evaporator thermo Compatibility must be at least equal with Compressor’s maximum high Rpm Btu Output. One problem with the BD80 compressor is it is a Low Back Pressure only unit, designed to operate at low temperatures between +23 degrees F and -22 degrees F, this maybe why Seafrost uses a low pressure regulator instead of a wide range +40 to -40 TXV.

Total heat absorbing service area of all three evaporators will determine their ability to remove Btu rate near that of BD80 compressor. Although asking for 100% efficiency is not realistic. To achieve good performance my guess is total surface area of plates exposed to conduct heat into plates should be 700 to 800 square inches. Less than this will increase cycling of compressor and possibly confuse AEO module’s control over performance

Refrigerant Flow Low pressure Regulator
The purpose of pressure regulator valve is to insure that a fixed pressure differential is maintained in evaporator so that the phase change from high liquid refrigerant inside these plates absorbs heat as low pressure turns it to vapor at a given temperature. The compressor is sucking low pressure gas vapor out of these plates as fast as it can while pressure regulating valve is holding plate suction to a pre adjusted value. If low pressure inside evaporator is regulated at 4 psi then incoming liquid 134a will vaporize lowering it to -5 degrees F. By turning hand knob CC wise on pressure regulating valve to 6.5 psi temperature in plate evaporator coil will increase to zero degrees F.

The problems with the above pressures and temperatures they are theoretical and are difficult to reach in these one of a kind system with multiple plate coils. . Again there are never two boats the same and in this case the low pressure gauge reading you would be dealing with to adjust valve is not what it is in the first, second, or third plate or even after the third plate as it is located at compressor. Pressure loss as refrigerant flows through plates and returning to compressor will cause some performance problems. At some point there will be a valve setting that can improve performance But first refrigerant quantity and purity must be checked. Best if pressure flow regulator valve setting is provided by system manufacturer.

Thermostat
Thermostat is the most important and most misunderstood component in these small Danfoss BD compressor powered systems especially when pushing the system output to maximum performance. The more thermostat cycles compressor with an variable speed AEO module it can reduce System Coefficient Of Performance (SCOP). Most small system designs control evaporator plate temperatures by stopping compressor before evaporator is inefficiently over powered by compressor. There is a trend to think that controlling box temperature with thermostat set to a very low differential instead of balancing component efficiency by controlling evaporator temperature is some how more efficient. It is true evaporator temperature is always at least 20 colder than box temperature and differential evaporator temperature between compressor stopping and restarting again varies from 8 to 18 degrees F. A wide evaporator differential setting will normally box temperature variance of 2 to 4 degrees. I would think thermostat temperature sensing tube bulb end would be attached against freezer plate near top. . If the system was 100% efficient and we know this one is not I would not use an area thermostat to control compressor cycling.


I believe poor performance trouble on Skips SeaFrost ice box conversion Refrigerator Freezer is caused by one or more of the following:

• It sounds like compressor may be cycling too often causing compressor speed to drop this will continue until AEO starts advancing speed until it again starts reducing speed. If the thermostat can be bypassed to cause compressor to run at full speed several hours and box temperature drops to -14 degrees then you will know if it is a thermostat or refrigerant flow causing poor performance.
• If performance is still bad with compressor running at max Rpm next step is to record high and low pressure readings every 30 minutes for 2 hours. This will eliminate condenser cooling as a problem. Also confirmation of pressure regulator performance.
• Contaminated Refrigerant ?
• ?

[Badsanta]

Richard, I wish that I knew half of what you do. Thanks for your concise explanations. I cant pretend to understand it all but you always make sense. Thanks for always helping.

[skipgundlach]

Hello, Richard,

Originally Posted by Richard Kollmann

Skips
Design


(clip...)

Total heat absorbing service area of all three
evaporators will determine their ability to remove Btu rate near that of
BD80 compressor. Although asking for 100% efficiency is not realistic. To
achieve good performance my guess is total surface area of plates exposed to
conduct heat into plates should be 700 to 800 square inches. Less than this
will increase cycling of compressor and possibly confuse AEO module's
control over performance

Aha. Perhaps this is the problem.

We have 3 11x16 plates, which yields only 528 square inches. 1/4 to 1/3
undersized.

However, as to cycling, while it's been a while since I tracked for days on
end, sitting looking at the lights every minute or so, my recollection has
the best at 2/3, but up to 4/5 of the time, the compressor was running. That
time usually was over 40 minutes at a minimum; typical off-cycle time was
10-15 minutes.

Is that too much cycling? If I'm on a set speed, would that matter?

Originally Posted by Richard Kollmann

Refrigerant Flow Low pressure Regulator

The purpose of pressure regulator valve is to
insure that a fixed pressure differential is maintained in evaporator so
that the phase change from high liquid refrigerant inside these plates
absorbs heat as low pressure turns it to vapor at a given temperature. The
compressor is sucking low pressure gas vapor out of these plates as fast as
it can while pressure regulating valve is holding plate suction to a pre
adjusted value. If low pressure inside evaporator is regulated at 4 psi then
incoming liquid 134a will vaporize lowering it to -5 degrees F. By turning
hand knob CC wise on pressure regulating valve to 6.5 psi temperature in
plate evaporator coil will increase to zero degrees F.

Hm. I know Clay spent considerable time getting the regulator, pressures
and volume to where he was happy with it. He initially set it at 2psi, but
then went up to 4psi. Indeed, Cleave instructed me to turn the valve CW to
drive the temperature down to where it would freeze the entire 3rd plate.

Not knowing where look, where would I find the temperature -5? With the
probe (per Sea Frost instructions) on the bottom of the last plate, we've
never actually achieved +2.5, full speed, non-stop. However, the box was
colder than we needed for even convective spillover, with the reefer going
below our normal cutoff point.

Originally Posted by Richard Kollmann

The problems with the above pressures and
temperatures they are theoretical and are difficult to reach in these one of
a kind system with multiple plate coils. . Again there are never two boats
the same and in this case the low pressure gauge reading you would be
dealing with to adjust valve is not what it is in the first, second, or
third plate or even after the third plate as it is located at compressor.
Pressure loss as refrigerant flows through plates and returning to
compressor will cause some performance problems. At some point there will be
a valve setting that can improve performance But first refrigerant quantity
and purity must be checked. Best if pressure flow regulator valve setting is
provided by system manufacturer.

I believe that Clay moved it up to 4psi after consulting with Cleave. In
any event, I of course marked before I did that, for reference. Due to
overkill in the initial manipulation of the regulator, reversing overcold
(ice past the box, let alone the last plate), it wound up well CCW of what
it had been, yet everything remained iced.

Unfortunately, after this most recent defrost, I'm again having to fiddle
with it, as the bottom quarter of the third plate didn't frost over.

Is this (constant adjustment of the regulator due to change in behavior
between defrosts) normal?

Originally Posted by Richard Kollmann

Thermostat
Thermostat is the most important and most
misunderstood component in these small Danfoss BD compressor powered systems
especially when pushing the system output to maximum performance. The more
thermostat cycles compressor with an variable speed AEO module it can reduce
System Coefficient Of Performance (SCOP). Most small system designs control
evaporator plate temperatures by stopping compressor before evaporator is
inefficiently over powered by compressor. There is a trend to think that
controlling box temperature with thermostat set to a very low differential
instead of balancing component efficiency by controlling evaporator
temperature is some how more efficient. It is true evaporator temperature is
always at least 20 colder than box temperature and differential evaporator
temperature between compressor stopping and restarting again varies from 8
to 18 degrees F. A wide evaporator differential setting will normally box
temperature variance of 2 to 4 degrees. I would think thermostat temperature
sensing tube bulb end would be attached against freezer plate near top. . If
the system was 100% efficient and we know this one is not I would not use an
area thermostat to control compressor cycling.

I'm living with the setup and have become accustomed to seeing temperature
ranges which would have had me in a panic with the Frigoboat system (recall
it was box temp only), as not only are they relatively momentary, they're
6-8 degrees higher than true (based on IR gun and analog thermostat at
sensor bulb).

So, our 6.5 minute hysteresis is appropriate? Or should it be longer?

Originally Posted by Richard Kollmann

I believe poor performance trouble on Skips
SeaFrost ice box conversion Refrigerator Freezer is caused by one or more of
the following:

• It sounds like compressor may be cycling
  too often causing compressor speed to drop this will continue until AEO
  starts advancing speed until it again starts reducing speed. If the
  thermostat can be bypassed to cause compressor to run at full speed several
  hours and box temperature drops to -14 degrees then you will know if it is a
  thermostat or refrigerant flow causing poor performance.
  
  

• Can I achieve that by setting the cutoff temperature at an unachievable
  level, and run it on high for several hours? A box temperature of -14 would
  suggest a plate temperature of 30 or more below zero. Is that correct? I
  wouldn't bet the boat on it, because anything could happen, but I can't
  imagine our system achieving that. (We've never had the temperature to
  +2.5.)
  
  In that case, would it be refrigerant flow at fault? And if so, what could
  be done about it?
  
  As to frequency of cycling, it's about an 80 minute cycle, with 50 or so
  minutes on and 30 off at best, or longer in the front and less in the back
  ends. Is that too much cycling?
  
  
  

  Originally Posted by Richard Kollmann

  If performance is still bad with compressor
  running at max Rpm next step is to record high and low pressure readings
  every 30 minutes for 2 hours. This will eliminate condenser cooling as a
  problem. Also confirmation of pressure regulator performance.
  

  I've not had personal gauges on this yet, but when Clay was doing his thing
  (which was before the water cooling), he would lay the side-top cover over
  most of it, lead the hoses out one side, and drape a towel across the gap to
  help in air-cool air flow (which is sucked out at the side and ducted
  overboard). Would I still need to do that with water cooling working? If
  so, is it reasonable to rely on readings in a not-really-sealed box?
  
  

  Originally Posted by Richard Kollmann

  Contaminated Refrigerant ?
  

  How would I determine that? Or if, after all the above, that seems possible,
  would a lengthy evacuation and recharge suffice to remove whatever
  contaminant it might be?
  
  

  Originally Posted by Richard Kollmann

  ?

Thanks, as always.

L8R

Skip

[skipgundlach]

Added info, in case it's useful to you:

The pump is as specified by Sea Frost. Their site claims it can handle
another couple of systems, so I presume the flow from a 1/2" intake and 3/8"
output is sufficient, if the lines are clear, which is the case at the
moment.

This is from a tee to a 3/4" input which goes on to the salt water washdown,
now compromised as it has air in it from sucking all the water out of the
cooling line if we use that washdown.

Their literature shows one amp, but Cleave claims 0.7A on the pump.

The feed line (3/8" output) is probably 10' to the compressor box/cooling
line input. The hose is bigger than the pump nipple, so has a sleeve in it
to seal. The out-to-drain line is about 5' or so. Both of those are
half-inch, I believe.

Whatever it is, however, is required to get my system to our temperatures; I'll try tomorrow with no water; I do not expect our system to cycle without it.

At this moment, in relatively cool (not over 90) weather, I'm doing detailed
time tracking on our closed boxes, with the control set on medium. While the
spillover was running, the freezer (indicated box) temps continued to climb.
When the spillover fan cut off at its setpoint, the freezer temp declined by
2.5° before restarting at an indicated 20.0, indicating the mass flywheel of
the water filling the bottom of the box, and the plates themselves.

The reefer spillover restarted before the freezer reached its setpoint of
6.5 on the bottom of the third plate, at an indicated 17.8. That dropped to
17.0 in a minute...

Over 4 hours, there were 4 complete cycles of the freezer starting from a
brief opening of each box before starting the actual recordings.

In the meantime, the reefer was cycling, as well, 6 times. As this was a
closed box, and a very narrow (2°) hysteresis, based on box temps, that was
very closely maintained. Times were relatively consistent, with the shorter
times being the run, and the off being longer. Cycles: 34/10, 33/12,
22/10, 30/11, 29/12 and 22/13. Combined, that's 68 minutes of run time
against 170 minutes off, or a runtime of 28.6%

The freezer, however...

Cycles: 51/20, 35/19, 34/13 and 41/19 run/off. Most of the run starts were
during a reefer pull, which drove warm air into the freezer. Indicated
temperatures (box) were suspiciously high, but the system cycled reliably,
for an overall 161 minutes of runtime against 71 minutes of offtime, for a
runtime of 69%, well off Sea Frost's 52% target, but down to an average of
58 minutes per cycle from what used to be 80.

[Richard Kollmann]

I will have time tomorrow to review all the information and answer your questions.

[Richard Kollmann]

Skip, My answers to your last two posts

Evaporator surface area Question?

I would not think plate surface area of 528 sq inches instead of 700 sq inches is the problem, so rule this item out. This difference between design and actual should have very little effect and if it did with thermostat controlling plate temperature instead of box temperature will compensate for the difference in evaporator size.. My guess is your BD50 system had expectable performance with a plate smaller than 500 sq inches when my book design for a large box calls for 504 sq inch plate surface area.

Question of how an AEO speed controller is influencing your units performance?

I have little experance with Danfoss AEO modules. I was asked to assist with a design problem on refrigeration units for a company building complete refrigerators. This company had a temperature controlled testing room where we could test performance at temperatures of 105 degrees F. In the two days experimenting I know the AEO performs as advertised by finding the lowest Rpm where thermostat is still able to satisfy desired box temperature.
On BD80 AEO control modules I do not have my normal testing data. On your system I can only guess what the compressor running profile will look like.
When compressor is first turned on it runs at slow speed providing a soft start after exceeding minimum control speed compressor will reach minimum operating speed of 3250 Rpm. From minimum speed module will start keeping count of running time of compressor as controlled by thermostat and degrees of thermostat differential (Hysteresis) before restarting compressor.

Example of AEO control when starting with a warm box and compressor is running at 3250 Rpm and beginning to increase Rpm 9.4 rpm/min. If thermostat does not see evaporator temperature reach its preset point compressor will run increasing Rpm minute by minute until 4400 Rpm is reached in 48 minutes. If evaporator does not reach temperature thermostat is set for compressor will continue on at 4500 Rpm, But this should not happen regardless of how warm the box is.

I normal operation with evaporator temperature setting of say zero degrees F is going to be reached in less than 48 minutes and thermostat will stop compressor. While compressor is off the thermostats differential (Hysteresis) of say 12 degrees will when allow compressor to restart. Each time there is a restart Rpm now will be 300 Rpm less than what it was when thermostat stopped compressor. During the time compressor was off evaporator continued to absorb box’s heat and compressor starts and another cycle begins.

Compressor cycles continue to lower box temperature to a point where a cycle is too short and speed needs to be reduced. AEO keeping track of this last cycle will again reduces 300 Rpm . When box temperature remote thermometer reads desired temperature AEO will extend cycle time. As long as thermostat is controlling temperature AEO will continue to increase and decrease Rpm with every cycle of compressor. If preset evaporator temperature does not provide desired box temperature set thermostat to a colder position.

Question Why is it necessary to run compressor at a lower speed? Being able to change compressor speed achieves two basic energy savings, it balances most mix matches of components in refrigeration systems application and lowering compressor Rpm improves its Coefficient Of Performance as much as 50% in small boxes or in cool climates.

What about Refrigerant flow regulator adjustment? And 134a Refrigerant pressure chart?

Refrigerant Flow Regulators adjustment depend on pressure drop in all in evaporators not only controls flow but also evaporator temperature. The pressure adjusting assembly has reference numbers that assist in setting low pressure but evaporator and line pressure drop is not available when you can only read pressure at compressor service port. The volume of saturated refrigerant flow through evaporator is where cooling should occur. Opening regulating valve raising low pressure increasing flow but cooling occurs at a higher temperature as low pressure increases. There is always a relationship between refrigerant pressure and temperature. With 134a Refrigerant 9.8 inches of vacuum in evaporator could produce cooling at -30 degrees F but with a BD80 compressor this might produce only 5 Btu of cooling in one hour. With an evaporator low pressure of 6 psi -20 degrees F. Again you can not read pressure in evaporates plates on this system.
The Pressure chart you are looking for is in my book page 113 or on web at: http://www.pacificseabreeze.com/tech-library/heat-and-insulation/R134a-PT-Chart.htm



What throws a monkey wrench into the normal operation of an AEO in most boat icebox conversions is a 10 cubic ft box is pushing the envelope for even the BD80 in tropical conditions no mater how well it is insulated.

• Knowing temperatures of liquid refrigerant coming out of condenser and refrigerant high pressure will rule out concerns over of air and water cooling.
• I do not know where a figure of 52% compressor running time comes from I would consider a compressor running over even 50% of the time for boxes heat load or its speed needs to be increased. When I ask for compressor cycle times I need one complete cycle generally recorded after the first day so I know box temperature is stable. A complete cycle is from compressor start, How many minutes did it run, How many minutes was compressor stopped before it was restarted by thermostat.
• Because a fan cooled condenser is a better stable cooling medium than water. I am not interested in water pumps or water cooling of BD compressors at this time. If this system will not perform correctly with air cooled it will not improve with water unless water temperature is cooler than 80 degrees.
• The test of Setting thermostat to run compressor continuously will help to rule out AEO module and maybe lead to what is really wrong with performance.
• I assume when refrigerant comes out of flow valve it is going directly to the freezer plate. Temperatures of lines going in and out of each plate and frosted areas my help to develop a superheat profile.
• First you tackle the non destructive test before messing with refrigerant or design changes. Sight glass observations are also important to compare against other reading.

[skipgundlach]

Originally Posted by Richard Kollmann

Skip, My answers to your last two posts

Evaporator surface area Question?

I would not think plate surface area of 528 sq inches instead of 700 sq inches is the problem, so rule this item out. This difference between design and actual should have very little effect and if it did with thermostat controlling plate temperature instead of box temperature will compensate for the difference in evaporator size.. My guess is your BD50 system had expectable performance with a plate smaller than 500 sq inches when my book design for a large box calls for 504 sq inch plate surface area.

This is the plate I had with Frigoboat.

https://www.coastalclimatecontrol.co...lat-plate.html

As my current plates are 11x16, that's a considerable percentage larger. (528 vs 651 and change, almost 25% bigger, and almost to your 700.)

Combine that with the nature of the medium, I expect that the FB pulled heat much faster, but gained it in the same proportion, making for a close hysteresis being feasible (or, at least, it seemed to work well for us).

Originally Posted by Richard Kollmann

Question of how an AEO speed controller is influencing your units performance?

I have little experance with Danfoss AEO modules. I was asked to assist with a design problem on refrigeration units for a company building complete refrigerators. This company had a temperature controlled testing room where we could test performance at temperatures of 105 degrees F. In the two days experimenting I know the AEO performs as advertised by finding the lowest Rpm where thermostat is still able to satisfy desired box temperature.
On BD80 AEO control modules I do not have my normal testing data. On your system I can only guess what the compressor running profile will look like.
When compressor is first turned on it runs at slow speed providing a soft start after exceeding minimum control speed compressor will reach minimum operating speed of 3250 Rpm. From minimum speed module will start keeping count of running time of compressor as controlled by thermostat and degrees of thermostat differential (Hysteresis) before restarting compressor.

Example of AEO control when starting with a warm box and compressor is running at 3250 Rpm and beginning to increase Rpm 9.4 rpm/min. If thermostat does not see evaporator temperature reach its preset point compressor will run increasing Rpm minute by minute until 4400 Rpm is reached in 48 minutes. If evaporator does not reach temperature thermostat is set for compressor will continue on at 4500 Rpm, But this should not happen regardless of how warm the box is.

I normal operation with evaporator temperature setting of say zero degrees F is going to be reached in less than 48 minutes and thermostat will stop compressor. While compressor is off the thermostats differential (Hysteresis) of say 12 degrees will when allow compressor to restart. Each time there is a restart Rpm now will be 300 Rpm less than what it was when thermostat stopped compressor. During the time compressor was off evaporator continued to absorb box’s heat and compressor starts and another cycle begins.

Compressor cycles continue to lower box temperature to a point where a cycle is too short and speed needs to be reduced. AEO keeping track of this last cycle will again reduces 300 Rpm . When box temperature remote thermometer reads desired temperature AEO will extend cycle time. As long as thermostat is controlling temperature AEO will continue to increase and decrease Rpm with every cycle of compressor. If preset evaporator temperature does not provide desired box temperature set thermostat to a colder position.

Question Why is it necessary to run compressor at a lower speed? Being able to change compressor speed achieves two basic energy savings, it balances most mix matches of components in refrigeration systems application and lowering compressor Rpm improves its Coefficient Of Performance as much as 50% in small boxes or in cool climates.

I'm a bit confused here (not surprising - I find I confuse easily these days).

I get it that lower is better in energy costs.

I had assumed that there was an ideal percentage of runtime. My sense of things has it that it would actually be best if the least rpm necessary to maintain the temperature were to run continuously.

Is it somehow more efficient to let a plate rise significantly, and then have to pull it down, than to keep it at a temperature (a big balancing act, I know, but perhaps achieving 85-90% runtimes would be better than say, 52% as Sea Frost wants it to get to)?

Originally Posted by Richard Kollmann

What about Refrigerant flow regulator adjustment? And 134a Refrigerant pressure chart?

Refrigerant Flow Regulators adjustment depend on pressure drop in all in evaporators not only controls flow but also evaporator temperature. The pressure adjusting assembly has reference numbers that assist in setting low pressure but evaporator and line pressure drop is not available when you can only read pressure at compressor service port. The volume of saturated refrigerant flow through evaporator is where cooling should occur. Opening regulating valve raising low pressure increasing flow but cooling occurs at a higher temperature as low pressure increases. There is always a relationship between refrigerant pressure and temperature. With 134a Refrigerant 9.8 inches of vacuum in evaporator could produce cooling at -30 degrees F but with a BD80 compressor this might produce only 5 Btu of cooling in one hour. With an evaporator low pressure of 6 psi -20 degrees F. Again you can not read pressure in evaporates plates on this system.
The Pressure chart you are looking for is in my book page 113 or on web at: http://www.pacificseabreeze.com/tech-library/heat-and-insulation/R134a-PT-Chart.htm

Cool chart. You'll recall the times with the dying Frigoboat, when I'd see low-side pressures deep into the negative psi...

But, recall that Clay had initially set up for 2psi, raising it to (a reported - he was finished before we got back from our trip) 4psi on Cleave's recommendation. Of course, all readings were from the compressor connectors.

As my only current frame of reference (haven't cracked the box to put on gauges) is the performance of the plates, I am only guessing at the current psi - but it's had to be different each time it's been restarted, in order to get the frost level complete.

Is that normal?

Originally Posted by Richard Kollmann

What throws a monkey wrench into the normal operation of an AEO in most boat icebox conversions is a 10 cubic ft box is pushing the envelope for even the BD80 in tropical conditions no mater how well it is insulated.

• Knowing temperatures of liquid refrigerant coming out of condenser and refrigerant high pressure will rule out concerns over of air and water cooling.
• I do not know where a figure of 52% compressor running time comes from I would consider a compressor running over even 50% of the time for boxes heat load or its speed needs to be increased. When I ask for compressor cycle times I need one complete cycle generally recorded after the first day so I know box temperature is stable. A complete cycle is from compressor start, How many minutes did it run, How many minutes was compressor stopped before it was restarted by thermostat.
  
  

• My bad - I went and looked it up - on pages 13 and 14 - and found that the target was 32 minutes, not 52%.
  http://www.seafrost.com/pdfs/BD-TW/B...rwater-pcb.pdf
  
  
  My runtimes were not in the 32 minute range. From the above:
  
  Cycles: 51/20, 35/19, 34/13 and 41/19 run/off. Total cycle times 71, 54, 47 and 60.
  
  The closest I got to the Sea Frost ideal runtime was 34 minutes. I'm guessing that the runtime is a fixed unit, and the offtime is whatever it will be for the box involved.
  
  As I haven't recorded this on automatic or high, I don't know what the runtimes will be, nor the ultimate amp cost. Would you expect that I'd do better at a higher speed?
  
  

  Originally Posted by Richard Kollmann

  Because a fan cooled condenser is a better stable cooling medium than water. I am not interested in water pumps or water cooling of BD compressors at this time. If this system will not perform correctly with air cooled it will not improve with water unless water temperature is cooler than 80 degrees.
  

  Well, I'm in trouble, for sure.
  
  Here's what happened today:
  
  This morning, as soon as the first cycle of the reefer and freezer quit, I set
  the control to high, and turned off the water.
  
  After more than 3 hours, when the freezer never cycled, at the next reefer
  cycle, I turned the water back on.
  
  29 minutes and nearly no temperature (box) change, it cycled. My take on it
  is that the plates never were really getting cold enough, despite a constant
  high-speed air-cooled input. Once the water was added, despite the lack of
  box temperature change, the plates got cold enough for the probe to see the
  trigger temperature, and it shut off, even though the box had little to no
  change. 18 minutes later, it came on again...
  
  In this period, the reefer went through 4 complete cycles, longer than
  usual, as the incoming air wasn't as cold as it usually was. Runtimes were
  more on the order of 40% vs the previously recorded 25%.
  
  So: Air cooling on high won't keep up, let alone cycle. I'm sure, from
  prior experience, that if I were to leave this on for a full day, the box
  indicated temperatures would creep from the current high of 24 to over
  "freezing" and the reefer would slowly take longer and longer to cycle,
  eventually running full time.
  
  The amphour delta for high speed compressor without, and medium with, water
  cooling is strongly in favor of medium speed compressor and water (see link
  to Sea Frost page which includes draw info).
  
  Our water is about 92°. There's a contradiction in reality here.
  
  

  Originally Posted by Richard Kollmann

  The test of Setting thermostat to run compressor continuously will help to rule out AEO module and maybe lead to what is really wrong with performance.
  

  Is the above illuminating at all to this above, or did you mean for me to lower the plate cutoff temperature to force a full-time run scenario?
  
  And, if you mean to exclude water, we already know that it will run forever, losing ground slowly, on high.
  
  

  Originally Posted by Richard Kollmann

  
  

  I assume when refrigerant comes out of flow valve it is going directly to the freezer plate. Temperatures of lines going in and out of each plate and frosted areas my help to develop a superheat profile.
  

  I've not tried to read the various points, and, due to my execrable internet performance where the boat is moored, I have to do this offsite and off my computer, I don't have a picture of the installation to attach, but, in a matter of the turn needed to make the connection, yes, the valve is right next to the first plate. I'll get the line and plate temps when I get back to the boat again.
  
  
  

  Originally Posted by Richard Kollmann

  First you tackle the non destructive test before messing with refrigerant or design changes. Sight glass observations are also important to compare against other reading.
  
  

  
  
  I'll have a look at the sight glass at each stage of whatever changes or recordings I make and integrate those to the chart.
  
  

  Originally Posted by Richard Kollmann

Thanks, as always.

L8R

Skip

[skipgundlach]

I ran out of time to edit; this is the link to the page with the amp consumption, right near the bottom:

BD Series

[skipgundlach]

Just now, with the reefer off (setpoint reached), and indicated box temperature 21.3 with constant rising and falling bubbles in the sight glasa, using the IR gun; ambient indicated 80:

Box probe at start 12
Regulator and attached pipe -1
Plate 1 -1
Pipe 2 -5
Plate 2 -6
Pipe 3 -7
Plate 3 -7
Pipe 4 (exit, next to plate) 0
Box probe at end 20
Indicated temp 21.3

10 minutes later, cycled, indicated temp 19.1 (not at exact cutoff; I missed that, and reefer had just started, so likely somewhat higher than without reefer warmer air).

I discovered that my freezer lid gasketing was disturbed during my last few defrosts; frost is currently severe and another defrost plus gasketing will have to take place before I can return it to previous levels.

However, is this useful info?

Thanks.

L8R

Skip

[Richard Kollmann]

A 20 degree Delta T between plate skin temperature and box temp is good for refrigerator areas but too large for freezer plate to freezer box.

I am almost sure refrigeration end of system has reached an equilibrium between all components when these reported temperatures were taken. To get a plate temperature of zero temperature inside is generally 10 degrees colder.

I do not know if you can reduce the Delta T and break the equilibrium by adjustments but it can be done by circulating air in box directly over plates. There is always natural air tumbling designed into refrigerator freezers as cold air descends and warm air rises. In this case conduction heat transfer is not good enough. By placing a 20 Cfm 12 volt muffin fan to blow air from top of freezer down across evaporator plate you improve heat transfer by convection air flow. There is a test in book demonstrating a 6 degree improvement by just tumbling air but if air were directed across plate it might reduces box’s temperature enough to solve some of the hot climate troubles you are seeing.