Rinnai Persistent Cold Water Sandwich: Firmware upgrade or other troubleshooting?

[ColdWaterSandwich]

Hi folks,

I've got a professionally-installed Rinnai RUR199iN, which has worked pretty well since it was installed over 4 years ago, including a dedicated recirculation line around the basement (~50 feet or so total), set to comfort mode. It's properly vented, gets enough natural gas, and has a clear condensate drain line. I descale it with vinegar yearly, most recently just a month ago. My water is whole-house softened and has very little calcium in it.

However, I've always suffered from "cold water sandwich" intermittently, either when taking a shower or in some other circumstances. I've gradually started to monitor the situation to learn more about the causes (and potential solutions), and via putting a temp monitor on the outgoing hot water pipe have been able to follow trends in how it operates (I get just one reading a minute from a SensorPush temp monitor which has been extracted from the plastic shell and the sensor placed on the outside of the hot water pipe... this logs trends but is inherently laggy).

The problem has been worse recently, probably because of the extremely cold December/January we had here. The input water temperature is down to 49F, whereas usually I think it's well into the 50s.

The most common scenario is the following:

1. Someone is taking a shower; water is nice and warm and constant
2. Someone else at a sink uses hot water for something like washing hands or rinsing a dish
3. They turn off the sink hot water.
4. Hot water temperature at the shower plummets, taking several minutes (sometimes 5-6 mins) to return to normal

I've now got a camera pointed at the unit and the timer controller, and yesterday got good data to show what's happening: When someone briefly increases the water draw, then shuts it off, the Rinnai turns off the flame for anywhere from 5-30 seconds before restarting. This introduces a super cold sandwich of tap-cold water. With the camera I can watch realtime the flame indicator on the controller as well as the outgoing water temperature as measured by the Rinnai using the conditions view (parameter 2 on the down-button/on-off switch diagnostics).

So, here's what I see for each of the above steps:

1. Starting point the outgoing temp reads close to the temperature set point (I've got it at 125F), steady flame icon (assume Rinnai is measuring 1.7 GPM)
2. As soon as someone opens a sink hot water tap, outgoing water temp drops to about 100 or so (assume hot water draw is now ~2.4 GPM). If the sink water remains on for ~30 seconds the temp will start to rebound
3. Within a few seconds of the person turning off the sink tap, the flame icon turns off. Output temp quickly drops through the 90s into the 80s. Eventually (5-30 seconds, typically about 20 seconds or so) the flame icon comes back on and the temps start climbing again, but it can take up to 5 minutes to get back to the set point.
4. Meanwhile, the cold water sandwich has made the shower unpleasant...

I'm at a loss to understand why the system is programmed to turn off the flame when the hot water flow drops by a just a third or so, particularly given the output temp is below setpoint at the time. Even if the system is creating too much heat for the new water flow rate, the bypass valve should handle tempering, I would think.

My system doesn't have any internet connectivity or even a USB port as far as I'm aware; could there be some firmware update that might address this problem? Googling for it talks about an update for the wifi controller (which I don't have) but I haven't seen any talk about firmware updates for the actual heater.

Any other troubleshooting tips for this? I've watched all the useful diagnostic parameters (bypass valve, temperature, flow rate, etc) and can't see any suggestion of some part of the system not operating "properly" ... I think all the parts are operating ok, just the way the thing is programmed seems a bit wonky. I'm also surprised by how long it takes to restore full temperature output once the flame kicks back on.

I will call Rinnai this week, but also wanted to get input from the very knowledgeable folks here. I've been a lurker for a while! Looking forward to contributing now that I'm a member.

PS: The installer who did my system has since retired; I was his first Rinnai (but he had done lots of Naviens)

 

[breplum]

1. double check that you don't have a bypass cross connection that we see sometimes in shower valves that go bad (especially Kohler 304 series and some old KWC). to check, turn cold water off at WH, open any hot tap and make sure it depressurizes down to nothing.
2. if no bypass, then my only other guess comes from constant Navien (with recirc pump models) issue where the internal check valve gets stuck open causing havoc with maintaining temperature due to fluid dynamics getting out of whack.

 

[ColdWaterSandwich]

1. double check that you don't have a bypass cross connection that we see sometimes in shower valves that go bad (especially Kohler 304 series and some old KWC). to check, turn cold water off at WH, open any hot tap and make sure it depressurizes down to nothing.
2. if no bypass, then my only other guess comes from constant Navien (with recirc pump models) issue where the internal check valve gets stuck open causing havoc with maintaining temperature due to fluid dynamics getting out of whack.

Thanks for the reply.

There's no bypass cross connection in my system. If I turn off the the cold water just before the WH, then open any hot tap at a sink or bath, there's no pressure after the initial short spurt.

I can also rule out the check valve as a problem; when I hook up my flush system each year, if the check valve had failed water would come gushing out. The check valve is located on the connection where the recirculation line returns. It's just fine.

I've also looked into operation of the internal bypass valve that tempers potentially too-hot water; the Rinnai reports it moving in ways that make sense based on demand.

I'm quite certain my problem is that the system shuts the flame down when demand drops, then gradually evaluates whether or not it needs to turn it on. I wonder if there is a secret maintenance/configuration menu not referenced in the installation manual?

 

[Twowaxhack]

Call Rinnai tech support. You’ll need a digital manometer and a quality electric test meter. You’ll also need to be standing in front of the unit with your tools before calling.

 

[ColdWaterSandwich]

Call Rinnai tech support. You’ll need a digital manometer and a quality electric test meter. You’ll also need to be standing in front of the unit with your tools before calling.

Good suggestion. I will need to buy a manometer (to test gas pressure?). I have a electric test meter that can do voltage and resistance. Anything else needed?

I can call them up and get specs on what I will need, then plan to do the real call once I have everything.

 

[GReynolds929]

There's no buffer tank in the water heater. You need a mini tank heater as a buffer tank to prevent the cold water sandwich.

 

[ColdWaterSandwich]

There's no buffer tank in the water heater. You need a mini tank heater as a buffer tank to prevent the cold water sandwich.

Thanks for the reply. The system worked reasonably well for four years so I don't think I need a buffer tank. Something has changed recently leading to increased sensitivity to changes in hot water flow. See below for some visualizations of the problem.

When I got the tankless, I accepted a little bit of "CWS" risk in exchange for endless hot water. But, what's happening recently is a problem. I'm learning more by the day in preparation for calling Rinnai. I'm busy at work these days so haven't yet made that call.

I mentioned above I have a thermometer logging temperature data once per minute; this data goes back at least a year. Here are some typical graphs of the recorded outgoing pipe temperature trends during "shower events" at various times. Note, the sensor is just taped next to the pipe, not actually in the water flow, so it's just an approximation representing the temp trends. A reading of ~98-100F means the output from the WH is around the set temp of 125F.

Here is what the heater output looked like during a typical shower a year ago:
(you can see the recirculation cycling before and after the shower)

(during the shower the temp was quite steady; I'm not sure what the upswing at the end was about... I just picked a random day)

Over the summer I started to notice variations more often but they weren't terrible. Here is a typical shower in August. Notice a sudden drop of several degrees toward the end of the shower.


Typical random day in September:

You can see the deflections are worse.

By October I noticed the cold spots were taking longer to recover. Look how protracted this deflection is:



And today; this was an admittedly bad example but there are a fair number like this recently:

After the first few minutes the temp bounced up and down a lot. It sucked!

I've learned that the deflections occur when someone uses hot water elsewhere, then shuts off that additional demand.

I did some experiments this afternoon and got some interesting data supporting the hypothesis that the system is, for some reason, over-mixing cold water into the heat exchager-hot water. I will post the details this evening or tomorrow. Thanks for your continued suggestions, folks!

 

[Twowaxhack]

I’d buy a new one and see if the problem goes away. I’m rich though…..

 

[breplum]

Any possibility that your shower valve is wonky? I've encountered thermostatic shower valves that got into negative feedback response loops with tankless. Replacing the thermostatic cartridge worked at once.

 

[ColdWaterSandwich]

I’d buy a new one and see if the problem goes away. I’m rich though…..

Right. I'll call that plan B. I'm not rich...

 

[Twowaxhack]

So you’re not getting a trouble code ?

 

[ColdWaterSandwich]

Any possibility that your shower valve is wonky? I've encountered thermostatic shower valves that got into negative feedback response loops with tankless. Replacing the thermostatic cartridge worked at once.

It's a good thought, but I see the same pattern even if I'm using the laundry room slop sink (which has separate valves for hot and cold). I see the same pattern no matter which sink/shower is in use.

So, this afternoon when I was alone in the house and thus could guarantee controlled water usage circumstances, I started up the shower at its usual settings (valve about 2/3rds hot, which makes a comfortable temperature, drawing about 1.8gpm of hot water). I let it run for a while to verify it would run stably until I messed with conditions in a controlled fashion. Here's the trace with annotations about what was going on:



For each numbered region here's what was going on:

1. Immediately after starting the shower, I went down to the WH and put it into the mode where you can see the diagnostic values (hold down the down arrow, press on/off button). I goofed and actually turned off the heater, resulting in that first little hump That was user error.

2. Looking at the heater output temperature, it quickly went to 122-125 and was stable in that range. My taped on thermometer took a while to come up to temp, but the heater was putting out consistently good hot water.

3. I let it run for a long time while assuring no other water usage. For a solid half hour, no change in output temp as measured by the Rinnai, and the external thermometer trace validates things were good.

4. I did a few things that use COLD water to see if the change in overall system pressure, thus flow, would be measurable. During this time I watched the Rinnai's GPM flow rate and it didn't vary perceptibly. It stayed constant at 1.8 GPM. the external thermometer didn't shift much; I checked at the WH output temperature setting (#2) and it remained between 122-125 during these tests. Then I shifted the view to be the valve Bypass % (how much cold water tempering the hot water from the heat exchanger)

5. The system had been running at a bypass value of 9 for all this time while the output was stable. I then did a 20 seconds run of hot water from a vanity sink (about an incremental .7 GPM), then shut off the sink. The bypass value was stable while the sink was running but then went up to about 50%, the flame turns off for 18 seconds (!!!). The flame turns on, the bypass value seeks around a bit, settling in the 20s.

At this point I ran some additional tests quickly looking at temp in the heat exchanger during repeats of the sink hot tap short burst, as expected the exchanger output temp plummets quickly to very cold temps and the actual WH output drops into the 70s...

Clearly, the system is programmed to assume the system doesn't chill as quickly as it actually does with my ~50F input water temperature, and reacts as if it is worried about sending scalding water through the lines.

 

[ColdWaterSandwich]

So you’re not getting a trouble code ?

No codes.

 

[Twowaxhack]

No codes.

Have you tried to force the heater to show codes ?

 

[ColdWaterSandwich]

Have you tried to force the heater to show codes ?

Yup! Position 1 it has a code 11 but the remainder are blank. That code 11 likely dates to last week when I turned off the gas for a minute when I was doing some testing. Last code I saw was 2 years ago when I got air in the recirc after a flush (code G3).

How can I clear that code 11? Or will it expire after a set time.

 

[Mitchell-DIY-Guy]

Internal flow sensor…my guess is this part should be replaced.

 

[ColdWaterSandwich]

Oh I forgot to mention, I have a camera pointed at the heater and timer/controller so when the system is displaying real-time conditions, I can view them live and later play them back for additional review.

 

[ColdWaterSandwich]

Internal flow sensor…my guess is this part should be replaced.

I've tested the GPM sensor, and found it to be reasonably accurate and reliable for values above 0.5 GPM. If I put it into GPM display mode, and vary the hot water draw, it gives good values. I tested accuracy by filling a 1 gallon jug at a few flow rates and it worked well given the 0.1gpm resolution. Below 0.5 it's a bit wonky, but the heater doesn't even turn on at such low flow.

 

[Twowaxhack]

Call rinnai tech support

 

[Twowaxhack]

The Rinnai shouldn’t shut off if the shower is flowing at least the minimum just because more hot water was called for at another fixture then that demand was satisfied.

Might have brain trouble with the unit if it’s not throwing a trouble code.

I would turn the circulation pump off and isolate it with a stop valve and see if the unit operates correctly under that condition.