Heating Help

Gotta love

@ May 12, 2013 2:38 PM in Steam at Work

the dog dozing peacefully through all the racket.

Yes

@ May 5, 2013 5:38 PM in More on Power Gas Burner Conversions

Given the level of interest in power gas burners you'd think more boiler manufacturers would have done this by now.

Surge protection & UPS

@ April 26, 2013 5:14 PM in A Steam Odyssey (Part 2): Midco Low-NOx Burner

No surge suppressor yet, but will be adding one soon. No plans at this time for a UPS.

User Interface

@ April 26, 2013 6:10 AM in A Steam Odyssey (Part 2): Midco Low-NOx Burner

I'll close this part of the saga with some pictures of the user interface.

If you're ever in the Philadelphia area and would like to see the system in operation, send me a message through the forums or to mstayton (at) ieee.org, and we'll see if we can set something up.

Instrumentation and Controls

@ April 25, 2013 8:48 PM in A Steam Odyssey (Part 2): Midco Low-NOx Burner

The control system has two outdoor and two indoor temperature sensors. The pairs can be configured to use the first good sensor, a specific sensor, or the highest, lowest, or average of the two values. The system also fetches the wind-chill temperature from a local WeatherBug.com weather station over the internet; if the wind-chill temp is lower than the outdoor sensor value, the system will use the wind-chill temp in the outdoor reset equations.

On the boiler there are temperature sensors for the tankless coil and the DHW temperature after the mixing valve, a type K thermocouple in the stack, and a 0-3 PSI, 4-20 ma pressure transmitter on the pigtail.

Off-boiler there is a temperature sensor on the riser to the farthest radiator, and temperature switches on the steam header and the (newly divided thanks to JStar) large and small mains. There's also a flow switch in the DHW line that will heat water on demand (with a 3-4 minute heatup delay if the boiler's cold) during scheduled DHW "off" periods.

I had good results with the Phidgets line of sensors on the previous version of this system, and selected their PhidgetSBC2 single-board computer as the platform to host the boiler control software. The SBC2 has an ARM processor and runs Debian Linux as the operating system. The sensors, control outputs, and control software are all on or interfaced directly with the SBC2; the system does not rely on a PC connection as was the case with the previous version. The user interface communicates with the the SBC2 over Ethernet on the home's local area network. While the SBC2 supports wireless, for this application I felt a wired Ethernet connection would be more reliable.

Safety & Backup Controls

The boiler is equipped with the usual safety features. On the 120 VAC side there's the Pressuretrol, LWCO, auto-feeder, and emergency stop switches at the boiler & in the cellar stairwell. During the conversion, JStar installed a flame rollout safety in the 24 VAC circuit to the Fenwal ignition control unit in the Midco cabinet.

Because this is still in the experimental stages, I've included some additional backup features.

There's a "Control Mode" switch on the front panel that selects between the Outdoor Reset/Modulating control, and the wall thermostat. In the "Thermostat" mode, a fixed voltage is supplied to the LNB burner. The control mode can also be selected from the user interface. And in the (hopefully unlikely) event of a computer or program failure, all of the control outputs are automatically set and held at their OFF state; this forces the system into Thermostat mode, and power must be cycled on the control box to return the system to modulating control.

A vaporstat is installed on the pigtail, wired to make on cut-out and break on cut-in. It is set to cut out at about 3 ounces of pressure and to cut in at slight negative pressure. On cut-out, an interval time delay relay is activated that cuts power to the burner for 30 minutes. This was added in the previous version of the control software to eliminate end-of-cycle short cycling on pressure, the idea being that if you're building pressure in the boiler then the radiators are probably full of steam and continuing to run the boiler is just a waste of fuel. Now that the new system is controlling on pressure, in theory the vaporstat should never activate, but I opted to keep it in for when the system is running in the fixed-output "Thermostat" mode. This limit control is always active and operates independently of the boiler control software.

There's also an emergency stop switch at the control box that removes power from the solid state relay that switches 120 VAC to the burner.

No webcam

@ April 24, 2013 10:21 PM in A Steam Odyssey (Part 2): Midco Low-NOx Burner

Sorry Eastman, no webcam. Not much to see on the outside anyway, all the action is in the firebox and inside the boiler. Neither of which are conditions that the average Best Buy webcam would survive for long. Would be great to see, though.

No word yet from the Midco folks.

Control Software Features

@ April 24, 2013 6:40 AM in A Steam Odyssey (Part 2): Midco Low-NOx Burner

Here are some of the key features of the control software:

1. Outdoor Reset Control using either time-proportional or heat loss mathematical models.
   
2. Supports simple on/off burner control and the following types of modulating control: Fixed Voltage, Modulating on pressure only, Modulating with low temperature override. (Based on early results with the Midco burner I will be revising the control algorithms to allow more flexiblity in the modulating control modes)
   
3. 7 day, 4 period-per-day temperature setback scheduling.
   
4. 7 day, 4 period-per-day Domestic Hot Water heating schedule.
   
5. Redundant indoor and outdoor temperature sensors, with configurable sensor selection (High/Low/Average/First Good).
   
6. Automatic switchover to a good sensor if one of the indoor or outdoor sensors fails.
   
7. Optionally retrieves current local wind-chill temperature over the Internet, and uses it as the outdoor temperature in outdoor reset calculations.
   
8. In non-modulating mode, eliminates short-cycling on pressure by entering a pressure lockout mode if an adjustable high pressure limit is reached.
   
9. On-demand DHW by pushbutton or when hot water flow is detected.
   
10. Highly configurable – many parameters can be adjusted to tune the system to its environment.
    
11. Logging of continuous analog data and heating cycle events, with trend charting for analog data.
    
12. Logging of fuel consumption data for the preheat, heat-on, and DHW cycles.
    
13. Automatically reverts to wall thermostat control in the event of a critical failure, or if selected by the user via a switch on the control panel. With the Midco burner, switching to "Thermostat" mode changes the burner control to a fixed voltage.
    
14. The user interface is written in Java and will run on any Java-enabled computer - Windows, Mac, or Linux.

Liner

@ April 23, 2013 6:53 AM in A Steam Odyssey (Part 2): Midco Low-NOx Burner

The liner is just on the back wall and the bottom, correct? Nothing on the sides?

Chamber liner

@ April 23, 2013 6:05 AM in A Steam Odyssey (Part 2): Midco Low-NOx Burner

Yes, there's a liner in the firebox. Isn't it supposed to protect the castings from flame impingement?

I'm willing to try removing it, but what's the risk? How do I know if we've done the right thing?

Thanks Dave!

@ April 22, 2013 7:46 PM in A Steam Odyssey (Part 2): Midco Low-NOx Burner

This has been an incredibly rewarding project, though not without risk. I usually prefer to avoid the "bleeding edge" and let somebody else prove a new technology first, but as a controls guy this was too good an opportunity to pass up. Heck, even coal boilers a hundred or more years ago had mechanical mechanisms to regulate them.

As with any new applied technology there are a few kinks to work out of the system, both the Midco and my controls.

Stack temps are still too high, and the Midco folks are working on that. I'm told they have several boilers in their lab, and I'm trying to find out what they're using that's closest to the firing rate of mine. Incidentally, the guys at Midco have been great to work with and are committed to see this succeed.

On the controls side, it really looks like the temperature override control won't be needed at all given the stack temps we're seeing. It's an opportunity to revisit the current control strategy and add new features that will allow even more flexibility for experimentation.

Payback's a

@ April 22, 2013 7:02 AM in A Steam Odyssey (Part 2): Midco Low-NOx Burner

wonderful thing. I know we don't talk price here, but did want to provide some idea of the expected payback period since the initial outlay is pretty hefty.

I prepay for 1100 gallons of fuel oil each year. Around these parts, natural gas (including supply, distribution, and other surcharges) is less than 40% of last year's prepaid cost of oil. That'll save me an estimated $2500 a year at current prices. Counting the cost of the burner, installation (excluding the piping rework), and the electronic & hardware components for the modulating control, the payback is less than three years. Maybe even sooner if you figure a modulating burner ought to use less fuel than a single-stage burner under similar conditions.

Incidentally, this made it a relatively easy sell to She Who Must Be Obeyed, although what really got her attention was the "greener" aspect of lower emissions from the Midco Low-NOx burner.

They'd save

@ April 22, 2013 6:03 AM in EZ-Gas Stays, Boiler is Upgraded

more fuel and money if you replaced it with a SafGard. I found that it took a couple of minutes to rebuild steam pressure after each time the CycleGard ran its intermittent level test during the heat cycle. But given what happened to the old boiler maybe the CycleGard will provide your client with more peace of mind.

Boilerpro's article

@ April 22, 2013 5:46 AM in A Steam Odyssey (Part 2): Midco Low-NOx Burner

Boilerpro wrote an article in 2009 with an interesting perspective on sizing steam boilers and venting. Here's the link:

http://www.heatinghelp.com/article/323/Boilers/1551/Taking-Another-Look-at-Steam-Boiler-Sizing-Methods-by-Dave-Boilerpro-Bunnell

That's a plan

@ April 21, 2013 10:41 PM in A Steam Odyssey (Part 2): Midco Low-NOx Burner

That's in my plans for the fall, I've been fascinated by Boilerpro's theory and would love to try it for myself.

JStar had an idea about running the system like a coal boiler; run the burner on "simmer" to keep the mains warm and then ramp up when there's a call for heat.

I built the controls to be flexible and configurable just so I could do experiments of this sort.

Limits

@ April 21, 2013 10:36 PM in A Steam Odyssey (Part 2): Midco Low-NOx Burner

The low fire output of the burner is 120 MBH (the specs say 100, but we clocked 120 on the gas meter). That's controlled by the RTC electronics in the Midco cabinet. It's probably limited by the ability to maintain a good fuel/air mix at lower blower speeds. I'd guess the limiting factor is the gas regulator; those electronically controlled motors can be driven accurately over virtually their entire range.

On Thursday we ran at low fire for 45 minutes while we did some initial testing. In that time we raised a head of steam, heated the mains, and started heating a few of the rads. One of the experiments for this fall when the weather is colder is to see just how low I can run the steam pressure while still getting adequate performance from the radiators.

Sort of

@ April 21, 2013 10:24 PM in A Steam Odyssey (Part 2): Midco Low-NOx Burner

The outdoor reset model I've developed knows the heat loss of the building and the heating capacity of the boiler. During a cycle, when the system transitions from Preheat (heating the mains) to Heat On, the software fetches the current wind chill temperature from a local WeatherBug station over the internet. The model uses the lower of the wind chill and outdoor sensor values in the equations that calculate the btu input and run time required to achieve setpoint. Lower temperature = greater heat loss = longer heat-on cycle time.

So while the software adapts to current conditions, it doesn't really learn from one cycle to the next.

Depends

@ April 21, 2013 10:08 PM in A Steam Odyssey (Part 2): Midco Low-NOx Burner

on how long the system runs. During recovery from setback (4 degF) most of the rads will be hot all the way across. On a normal heat cycle maybe half to 2/3 of the way across. In mild weather it's even less than that.

It'll be a while before I have enough data to tell what difference the burner makes. What I know right now is that I'm controlling the burner at just under an ounce of pressure and am getting steam to the rads, where with the oil burner it was at 1.75 ounces. That's got to be saving fuel and $$$. But I think that's more a function of the corrected near-boiler piping than the burner.

Heat Loss

@ April 21, 2013 9:28 PM in A Steam Odyssey (Part 2): Midco Low-NOx Burner

I used the Slant-Fin Heat Loss Express software to calculate the heat loss for the outdoor reset model. It's 133 MBH if you count the 3rd floor (used for storage; there are two radiators up there, one's turned off), and 106 MBH for just the 1st & 2nd floors. The model has been more accurate at predicting the heat-on time using the lower number.

That looks

@ April 21, 2013 9:11 PM in EZ-Gas Stays, Boiler is Upgraded

a lot like my Utica. Can't imagine why. :)

Is that a CycleGard or a SafeGard LWCO?

Stack temps

@ April 21, 2013 6:57 PM in A Steam Odyssey (Part 2): Midco Low-NOx Burner

with the oil burner were between 465 and 500 degF. Note that my thermocouple is in the flue pipe about 30" from the boiler. Combustion analyzers probes are inserted much closer to the boiler and read higher temps..

This burner

@ April 21, 2013 6:52 PM in A Steam Odyssey (Part 2): Midco Low-NOx Burner

is right up against the Kaowool liner in the firebox.

Real world conditions

@ April 21, 2013 8:32 AM in A Steam Odyssey (Part 2): Midco Low-NOx Burner

Overnight temperatures were in the 30's so I got to see the burner perform under real world conditions. Here are the statistics for two runs this morning. The Preheat cycle is the time it takes to heat the mains and see steam at the condensate sensor on the farthest radiator.

First run
Preheat: 8:42 / 35.0 MBtu
Heat On: 31:13 / 122.4 MBtu
Total Cycle: 39:55 / 157.4 MBtu

Second run
Preheat: 7:42 / 31 MBtu
Heat On: 19:49 / 75.4 MBtu
Total Cycle: 27:31 / 106.4 MBtu

Both Runs
Preheat: 16:24 / 66.6 MBtu
Heat On: 51:02 / 197.8 MBtu
Total: 1:07:26 / 264.4 MBtu

The Beckett oil burner was rated at 245 MBH. For the same total time, that's 273.6 MBtu. There's not enough data yet to tell if the difference is significant.

I've attached a plot of the two runs. I need to spend some more time tuning the pressure controller to eliminate the oscillation when the pressure first comes up.