Joined on November 14, 2009
Last Post on August 1, 2014
@ December 1, 2013 1:03 PM in Midco Low-Nox Burner Performance ReportMaybe I'm mis-using the term. In my system, there's a differential and an anticipator value. The differential is the amount below temperature setpoint that will start a heat cycle. The anticipator is the amount below temperature setpoint that will shut off the burner ("anticipating" that the temperature will coast to or past setpoint). I usually keep them within 0.1 or 0.2 of each other.
@ December 1, 2013 12:56 PM in Midco Low-Nox Burner Performance ReportHi Rod - Two things might help: insulating the returns and bringing in outside air for combustion.
I have almost 100 feet of return piping, some of which is behind a false wall, not to mention that much 1-1/4" insulation is around $200. Unless someone has some hard data, I don't see the return on investment.
Outside combustion air is probably the better bet, since it should decrease stack temps and increase efficiency, which in turn might cut preheat times. To do it right I need to break through two 1-foot thick stone walls. I might try rigging something temporary through a window to test it, but that's still almost 25 feet from the burner. Plus I wouldn't be able to tell anything without a combustion analyzer, so I'd need to wait until JStar was in the area again, unless I can locate the one that our combustion controls group at work is rumored to have.
One operational change that is having a big impact on preheat times is an experiment I'm currently running where the heating cycle times are reduced 50-66 percent (roughly 2-3 CPH). A nice side effect of this is that the pipes are still hot when the next cycle starts, so preheat times are down to just a couple of minutes. The sample size is still too small to draw any conclusions about fuel economy, but once I have more data I'll post a report.
@ December 1, 2013 11:59 AM in Midco Low-Nox Burner Performance ReportSensors everywhere except for snap disc thermal switches on the steam header and the exits from the steam takeoffs at the mains. Temperature sensors are all 10K thermistors. Pressure sensor is a 0-3 psi, 4-20 ma transmitter. Stack temperature is a type K thermocouple.
All sensor data is logged and time stamped at least every five seconds, more often if the data is changing rapidly such as when the burner first fires. Event data such as a cycle change, temperature switch closure, etc . is logged, time stamped, and marked with a unique heat cycle ID.
@ December 1, 2013 11:29 AM in Midco Low-Nox Burner Performance ReportPerformance Summary for Oct 1 through Nov 30, 2013:
Number of Heat Cycles: 274
Total Run Time (hrs:mins): 81:08
MBtu Consumed: 15,662 (21% less)
Degree Days: 865
Fuel Cost: $154.22 (70% less)
Btu per Degree Day per Square Foot: 6.02 (32% less)
Equivalent Oil MBtu Consumed: 20,203
Unit Cost of Gas vs Oil per MMBtu: 62% less
Oil Btu/DD/sf from 12/2003 thru 4/2013: 8.80
Preheat Cycle Time (min:sec): 05:14
Heating Cycle Time (min:sec): 12:32
Total Cycle Time (min:sec): 17:46
Preheat MBH: 239 (2% less)
Heating Cycle MBH: 187 (24% less)
Total Cycle MBH: 193 (21% less)
1. Percentages are gas compared to the single-stage oil-fired burner.
2. Oil MBtu consumed is based on an oil burner firing rate of 245 MBH for the same run time.
3. Unit costs for fuel are based on average as-delivered costs in the Philadelphia area for Oct/Nov 2013
@ December 1, 2013 11:26 AM in Midco Low-Nox Burner Performance ReportFirst in a series of reports on the performance of the Midco LNB-250 modulating gas burner on a residential one-pipe steam system.
Boiler: Utica Starfire III SFE, 152 MBH Net
New Burner: Midco LNB-250 Low NOx modulating gas burner, 75-300 MBH
Old Burner: Beckett 245 MBH single-stage oil burner
Radiator EDR: 607 sq ft
Building Living Space: 3009 sq ft
Click here for more details
System Operating Characteristics:
The overall heat cycle is divided into two sub-cycles: Preheat and Heating. The Heating cycle is further divided into Heat On and Heat Off segments. A new heat cycle begins when indoor temperature is below the comfort setpoint.
Two snap disc temperature switches are used to indicate when steam has reached the drop header and has entered the mains. The switch contacts close at 160 degF and open at 150 degF. An analog temperature sensor is used to indicate when steam has filled the mains and the Preheat cycle ends.
Two PID velocity algorithms modulate the Midco burner under either pressure or stack temperature control.The pressure control setpoint is 0.8 oz/sq in. The stack temperature control setpoint is 670 degF.
At startup, with no pressure in the system, the burner output increases until the stack temperature is reached. The burner output then modulates to maintain the stack temperature until pressure builds. Once pressure reaches setpoint, the burner output modulates to maintain pressure. Depending on the length of the heat cycle, the burner output gradually decreases from 240 MBH to 78 MBH.
Typical Preheat times from a cold start (mins:secs):
--> Burner On to Header Hot: 07:00
--> Header Hot to Main Hot: 02:30
--> Main Hot to Preheat Sensor Hot: 02:45
--> Total Preheat Time: 12:15
The length of the Heating segment is determined by a heat-loss replacement algorithm that calculates the btu input required from the burner to achieve the indoor temperature setpoint. The burner runs for the calculated time or until the indoor temperature setpoint (minus an anticipator value) is reached, at which point the burner is turned off and the system enters the Heat Off segment.
The length of the Heat Off segment is the difference between the Heating segment and the overall Heat Cycle time. After the Heat Off time, the system enters the Idle state and is ready to run another heat cycle.
@ November 30, 2013 9:19 AM in Hydrolevel VXT alternative installThe way you have it laid out, you won't be able to feed water to the boiler if the VXT needs to be removed for some reason.
@ November 28, 2013 9:43 AM in New steam boiler0-3 psi gauge
If you're running less than 1 psi you can substitute one of these:
0-15 in/H2O (about 9 oz)
0-30 in/H2O (about 1.1 psi)
@ November 25, 2013 5:26 PM in Weil-McLain EGH ratings loweredHi Rod, just as a point of interest my 2009 Utica SF does have pins all the way up the sections, and they're shown that way in the I&O manual. The T baffles that JStar installed rest on the pins of each section. I think he took a picture, maybe he'll post it.
And if Burnham/Crown wants to donate a free MegaSteam or FSZ block I'd be happy to take it. :)
@ November 24, 2013 2:53 PM in Gorton Water FountainsAre the vents on an "antler" 6" or more above the main? If not and you have the room I'd do that first.
@ November 24, 2013 11:13 AM in Two Men and a Burner (A Midco saga)The LNB-500 burner is rated at 229 cfm. Mine has an orifice plate to bring the max btus down to 300 MBH, so if it's a linear relationship the max cfm is 137. I never run at more than 80% of full speed since my btu input requirements are lower, so again if it's linear that's 110 cfm.
How do you keep critters, leaves, bugs, water, etc. out of the intake pipe?
@ November 24, 2013 10:25 AM in Two Men and a Burner (A Midco saga)I'm sure it would help. It'son the list of things to try, but it involves breaking through two 1-foot thick stone walls to bring in outside air.
@ November 24, 2013 1:14 AM in Two Men and a Burner (A Midco saga)... I would ditch the Utica boiler, buy the Midco and put it on a MegaSteam (or the Crown FSZ equivalent). Because I firmly believe that most of the issues we've been working through have more to do with the boiler than the burner. On the Utica, it's a straight shot from the burner 26" up through the sections to the stack. I think the out-of-the-box stack temps and efficiency would be much better and require much less, if any, tweaking with a Midco on a 3-pass boiler like the MegaSteam.
This all started in 2010 as trying to solve the problem of end-of-cycle short cycling on pressure. I do industrial process control software by day, and knew that the "right" solution was to adjust the burner to maintain an even pressure. Problem was, there weren't burners that could do that in my btu range, until the Midco came along. And so far as I know, there still isn't a commercially available pressure control that's plug-and-play for this. Ah, the joys of adapting 21st century automation to 19th century technology.
Let me add that I'm neither discouraged nor disappointed by the experiences thus far. No one knew exactly how the burner would perform on this particular boiler. At some point you need to get these units out of the lab and onto real-world systems. Early adopters of any new technology need to expect a few bumps along the way, to smooth the road for those that follow.
The performance data is telling me it's all worth it.
@ November 23, 2013 9:46 PM in Two Men and a Burner (A Midco saga)Last month JStar came out to help diagnose a hard light-off issue I'd been occasionally experiencing with the Midco burner. Long story short, we ended up rotating the burner head so that the spark rod assembly was on top of the burner, rather than below as supplied from the factory. Once we could watch the spark thru the firebox viewport (couldn't see it when it was underneath the burner), we observed that the spark would jump between the ground rod and the burner mesh. After repositioning and re-gapping the spark rod, we saw good ignition every time. I sent a detailed report to Midco describing our observations and changes, and they agreed with what we'd done. In over 200 firings since, there have been no hard light-offs.
In August, when Midco replaced the 500 MBH burner head with a 300 MBH version they made up custom for this boiler, stack temps were still running high at 780 degrees at high fire (260 MBH), efficiency was around 72% (78% at low fire 75 MBH), and excess air was 42%. It was pretty clear that a lot of heat was being lost up the stack, but there wasn't much else that could be done to the burner to fix it.
Today, JStar brought some ideas to try to bring down the stack temperature and increase the efficiency. First he installed some angle iron T baffles between the boiler sections. On the very next firing we saw the stack temperature drop by 100 degrees and the high fire efficiency increase to 78%, with excess air at 34%.
The Midco has a variable speed blower so there's no other adjustment for combustion air. Joe brought in a HUGE PVC ball valve and attached it to the blower inlet. We tried several valve positions to restrict the combustion air to the blower and took analyzer readings for each. With the valve 5/8ths closed, Joe got excess air down to 11.5% with 81.1% efficiency at low fire, but CO went through the roof. The "sweet spot" seems to be with the valve half closed, with the following readings at low and high fire: stack temp = 575 / 685, excess air = 19 / 26.5, efficiency = 80.6 / 78, CO = 25 / 8. If you're interested, here's a link to all of the combustion results: Combustion Results
In addition to the combustion analyzer readings, I was also watching the stack temp and steam pressure readings on my control system. With 1 ounce of pressure at the boiler and a constant firing rate, we saw steam pressure increase almost 2/10ths of an ounce when we introduced the 50% combustion air restriction, and half an ounce when 5/8 restricted. At first I thought it was a fluke but the results were repeatable, a clear demonstration of the actual impact changes like this make to system performance. Pretty neat.
Next Sunday I'll be posting a summary report of the Midco's performance data for October and November.
And finally, a picture of the reconfigured Midco burner with the combustion air restrictor valve in place. Kids, don't try this at home.
@ November 22, 2013 10:18 PM in Homemade ControllerI think you're right about not scaling pressure. 2 ounces is 2 ounces regardless of whether it's a 1/4" tube or a 3" pipe. I think you'd need to scale the boiler water content and btu input to match the EDR of your test system.
I applaud your scientific spirit and look forward to hearing about your results. For what it's worth, I ran a lot of simulation testing in the six weeks between completing my control software and when the new burner was installed, and it paid off. I'll admit to a few white knuckle, is-this-really-gonna-work moments before it came time to throw the switch, but in the end all I really had to do was adjust the tuning parameters on the pressure control to suit the real-world response.
As somebody once alliterated, prior planning prevents poor performance.
@ November 21, 2013 8:02 PM in Homemade ControllerThose numbers seem pretty good, especially running that big garage rad and a boiler that old. The construction here is similar - 1890-ish four course brick with a 2" air gap between brick and interior wall on 1st & 2nd floors, 3rd floor & attic are insulated, low-e glass in windows. How closely matched is your boiler to your EDR? Mine's about 4% oversized so not a bad match.
@ November 21, 2013 4:26 PM in Homemade ControllerIf there's no internet connection the software will just use the wired outdoor temperature sensor values. The control platform itself is autonomous, it will run regardless of whether an Ethernet link is available or if the client user interface is running.
I've read here that the best method of comparing two different buildings is btus per degree day per square foot of floor space. If I remember correctly, under 5 is awesome, 5-10 is good, 10-15 is OK, above 15 is not so great. Substituting EDR for building square footage ought to work too.
Edited to add: If you change gas cfm to btus in your calculation, you'll normalize that variable so you can compare to systems burning different fuels. I just ran a btu/edr/deg day calc with my current Oct/Nov data and get 29.1. My btu/dd/sq ft is 5.88.
@ November 21, 2013 3:28 PM in Homemade ControllerWell, my system certainly falls in the "mad scientist" category, but that's intentional - it's built for flexibility, experimentation and data acquisition. Once I get a handle on the best combination of fuel economy and comfort I'll probably dial it back to a subset of the overall functionality. I'd have to eventually anyway, it's hard to sell a house with Franken-controls in the basement!
That little PLC of yours looks pretty good with a great price point. Might have to get one to play with. I see they have an analog output extender board which I'd need to modulate my burner. It looks like it has function block programmability in addition to traditional ladder logic, is that right?
Regarding heat loss, I compensate somewhat for wind by fetching current wind-chill temperature from a local weather station over the Internet. It will nudge the run times up if there's enough difference between outdoor sensor and wind-chill temps. There are still some variables I don't fully understand, so sometimes it'll miss the target (over or under) by a degree or so, but for the most part the math has been pretty good at getting within -0.2 to +0.7 degree of target temperature. Even so, it's really just a guess. Your "adaptive pulsed modulation" method could replace that.
Curious if you have an atmospheric or a power burner, and if you were able to check combustion efficiencies while you were manually modulating your gas valve.
@ November 21, 2013 1:28 PM in Homemade ControllerThis is a very interesting control strategy PMJ, I'd like to learn more details about it. You've got me curious how well your strategy might work on my system, one-pipe running 1 oz or less pressure (someone once called it "near-vapor") with a modulating burner. Right now, once the mains are hot the burner turns down to maintain a low pressure setpoint, and my average firing rate is currently 16% less than a fixed output burner. That's on top of the general cost savings of gas vs oil. I use a "heat loss replacement" model (think outdoor reset on steroids) to calculate run times. It works pretty well, but it would be nice to reduce the cycle-to-cycle heat variance in the rooms if I can maintain (or reduce) fuel usage at the same time. More cycles should mean less preheat time which is also less time at high fire.
@ November 21, 2013 6:09 AM in Homemade ControllerThanks. It's been a fun project. I've learned a lot about my own system and steam systems in general, and I'm saving fuel and $$$.
One recommendation I have for you is to make sure your system is tuned up - proper main venting, pipes insulated, etc. Controls can't fix a system that isn't already in good working order.
Stay tuned. Around Dec 1 or so I'll be posting a performance report for October & November. It's something the folks at Midco asked me to do, and I figured I share it here as well.
@ November 20, 2013 6:07 AM in Homemade ControllerCheck out this Phidgets project created by high school students from Phoenix AZ in a program that provides the students with hands-on experience in science, technology, engineering and mathematics (STEM).
Falcon Autonomous Underwater Vehicle
@ November 19, 2013 9:55 PM in Homemade ControllerI've been working on such a "homemade controller" for several years. I developed a heat loss model for calculating boiler run time, and more recently added PID pressure control when we swapped out the oil burner for a Midco LNB modulating gas burner back in April.
I chose Phidgets (www.phidgets.com) as the hardware/sensor/control platform primarily because you don't have to mess with low-level circuitry like you do with Arduino and can just get on with the job of programming it to do what you want.
Earlier versions of the control did have a "pressure lockout" function similar to what you describe, but these days it's more or less out of service since the new burner's firing rate is controlled by steam pressure. When it was in service, the "pressure lockout" worked quite well and saved more than a few gallons of oil over the course of a heating season.
The three links in the last line of my signature will take you to a couple of threads on the Wall, and an external web site where my project is described in detail.
@ November 11, 2013 7:24 AM in Pipes leaking at shutoff connectors (All of them)My old instrumentation is mostly pre-1950.
Incidentally, a modern Honeywell vaporstat can be set to cut-in on slight vacuum. I had mine set to cut out at 4 ounces pressure and cut in at about 1/2 to 1 ounces vacuum. You just set the differential to a higher value than the cut-in. I didn't test it below 1 oz vacuum though, so I don't know how accurate it would at higher vacuum.