In my last post, I discussed a technique one of my mentors taught me for assessing steam consumption using an alarm clock. A recent project provided an opportunity to demonstrate the validity of the technique by cross checking the data for steam consumption based on condensate pump cycles against data based on flow and differential temperature. The results are illustrated graphically below.

For the days illustrated in the graph, the only significant load on the steam system was the preheat coil on a constant volume air handling unit which, thanks to the efforts of Erich Blaufuss at 8760 Engineering, I knew the flow rate for and for which I had logged differential temperature at the preheat coil.

The other load was a constant volume reheat water system, which I had logged differential temperature data for along with a pump test and a pump curve, which established the flow rate.

In the graph, the green line, which is fairly flat around 1,500 lb. per hour with repetitive 200 – 300 lb. per hour spikes is the load calculated for the hot water system using the following equation.

Since the system is a constant volume system, I was able to determine the flow rate by a simple pump test. Then, I took the logged temperature data, exported it to a spreadsheet, and inserted a column to do the load calculation with the equation illustrated above.

Since the equation gave a result in Btu per hour and I was interested in pounds of steam per hour, I added a column that used the Btu per pound factor for the steam on the site to convert Btu per hour to pounds per hour and selected that column for the graph.

To determine consumption (which is not shown on the graph), I had to take the load, which was in lb per hour and integrate it over the logged interval. Since the logged interval was 2 minutes for the temperature data, that meant needed to add a column to my spreadsheet that:

- Took the pounds per hour I had calculated as described in the preceding paragraph,
- Divided it by 6o minutes per hour(giving me pounds per minute) and then,
- Multiplied it by 2 minutes (the logged interval)

which would give me the consumption for that particular interval. The consumption for any given time window then could be assessed by adding up the consumption for each logged interval between the starting and ending date and time of interest.

That process may sound a little intimidating if you are not accustomed to doing calculations like that and/or are not familiar with spreadsheets. But, its one of those things were it sounds worse than it is when you write it out. Towards that end, I’m thinking of doing a series of posts that walks through the details of different calculations, including screen shots of the spreadsheets and formulas, starting with the details of the calculations I used for this example so stay tuned.

Returning to the example, I assessed the preheat coil load via similar technique to what I just described for the reheat system, but used the air side sensible heat equation instead of the water side equation. Here is the air side equation as a memory jogger.

The flow rate was constant and furnished by Erich, as mentioned previously. The temperature rise came from data loggers, just like it did for the hot water system. Once I had the load, I added it to the hot water load and plotted the total, which is the blue line on the graph.

The condensate pump based load, which was calculated using the technique I discussed in my previous post is the red line that has a shape similar to the blue line, which is my point. The correlation between the two methodologies for measuring load is pretty good all things considered, at least based on my experience. And, its a lot more than we knew when we stared since there was not a working, independent steam meter for the building.

The differences between the two data streams are likely related to the accuracy of the instruments and the batch nature of the condensate pump monitoring technique. In other words, there is a steady flow into the condensate pumps, but the flow out, which is what I measure, is in “chunks” directly related to the receiver volume between the high and low float switch settings. Under heavy loads, their will be more “chunks” per unit time than under low load. If you look at the logger data, its variable frequency square wave that I converted to a load using techniques I will discuss in subsequent posts.

Meanwhile, back on the home front, I’m fiddling around (technical term) with automating and monitoring my my home energy consumption with a powerline carrier/wireless technology called Insteon using a Universal Devices ISY 99i. That’s the automation side.

The monitoring side, for now, will consist of a Brultech ECM 1240 for the electrical side. Gas is a little tricker. To do that, I’m using Mamac Maverick that will log flow with a Dwyer MS Series sensor using velocity pressure as a proxy for flow and log temperature using Omega 4-20 ma transmitters with RTDs . With that data, I can calculate the load using the air side equation I discussed above.

The Mamac solution is a bit “spendy” for a home, but in my case there are a couple added benefits. For one thing, if you’re like me, your house (and maybe the world) is just one big science experiment and this is kind of like a hobby. But from a practical standpoint, the experiment also lets me pilot test the Maverick’s web enabled access. If things work the way they are supposed to, I should be able to see it from where-ever I am when I travel and make it send regular CSV (Comma Separated Value) files to a couple of e-mail addresses. Assuming I can get it to work (and I think the limitation here is my knowledge with regard to IT type stuff like IP addresses, subnet masks, etc.) it’s going to become a low cost ($275 per controller plus what ever you spend for sensors) way to do remote monitoring, providing a way to get the data with out visiting the site; quite attractive when the site is in California or the Midwest or the East Coast and you are in Portland.

So, in the future, watch for more on those fronts too. Meanwhile, here’s to a happy holiday for all of you.

David Sellers

Senior Engineer – Facility Dynamics Engineering