Understanding an Anomaly – Part 3 – A Hypothesis

Performing an average daily energy consumption analysis for the lab facility I have been discussing in the past several posts resulted in the following for pattern for gas consumption.

The lack of gas consumption prior to February of 2006 along with the change in the electric consumption profile after that point in time clearly indicates when the new gas fired heating equipment came on line and the existing electrically driven heating equipment was decommissioned and shut down.

To check the correlation of energy consumption with climate, I decided to plot the average heating and cooling degree days for the lab location on the graphs. Weather data of this type for a specific location and date range can usually be found on line at the National Climate Data Center for a small fee.More general statistics – for instance average heating degree days for a specific location – can often be found on line at no cost from one of the regional climate centers.

Heating and cooling degree days generally indicate the need to heat or cool to maintain comfort conditions in a given climate. But, contrasting energy consumption for a complex building like a lab with degree-day data must be done with a “grain of salt” as the saying goes.

For one thing, the degree day calculations assume a balance point where the internal gains exactly equal the losses and no heating or cooling is required (typically, 65°F is assumed, but other base temperatures are also used). For a building that is well insulated and/or with significant internal gains, this may be a bad assumption.

In addition, degree days do not directly take the need to humidify or dehumidify into account, so they can be misleading if they are used to calculate energy consumption for cooling or heating. None-the-less, they are a good indicator for the need to heat or cool and can be useful for deciding if consumption patterns are climate driven or are driven by something else.

For the facility that I have been discussing, the gas consumption tends to follow the shape of the heating degree day data (as did the electrical consumption when electricity was the source for heat). In addition, 30% of the peak consumption seems to be related to base load, based on the consumption in the summer months.

A pattern like this often be an indicator of simultaneous heating and cooling. But, at the Oregon coast, night time temperatures can dip into the 50’s even in the summer (its realy nice, you should come visit – stay in Neskowin) and thus, there can be a need for some heat, even in the summer, especially for lab systems with high outdoor air flow requirements.

Here is what the climate looked like in 2006 based on local hourly weather information which can be obtained from the National Climate Data Center for a small fee .  Each spike is one day.

All of the preceding doesn’t mean that there is not simultaneous heating and cooling going on. Given that the facility is a lab with minimum flow rates fixed by air change requirements and discharge temperatures set by dehumidification requirements, there is probably a reheat load.

Complimenting these consumption based observations is a field observation I made while I was on site on a mild day. Specifically, I notice that a boiler would cycle on and fire for 5-10 minutes and then cycle off for 2-3 minutes and then repeat the cycle.

If the cycling can be minimized via set point optimization or some other strategy, the potential exists to improve efficiency by eliminating the losses that occur during the boiler’s pre and post firing purge cycles. Thus, my somewhat casual field observation complimented by my knowledge of the general nature of laboratory facilities, the climate at the Oregon coast, and the average daily gas consumption pattern at the facility indicate that there is probably some potential for reducing the baseline consumption.  But, they also say that I probably can’t eliminate it.

In contrast, the current electrical consumption pattern has little or no correlation to the climate data.

While the need to cool in the mild Oregon coastal climate is minimal, there are still days with temperatures in the 80’s and even the 90’s°F (very nice if you are out on the beach, especially with a sort of sweet white wine) in the peak summer months.

Yet, these extremes seem to have little impact on the energy consumption pattern at the facility, even though it the HVAC systems there handle significant quantities of outdoor air. Discussions with the operating staff revealed that in fact, the chilled water system does not operate that frequently and that outdoor air can be used to provide the necessary cooling much of the time.

To me, this said that while there may be opportunities to optimize the cooling equipment, they probably should not be my first focus. Stated another way, if I could magically keep the facility comfortable with out ever operating the chilled water system, I probably would not have much effect on the over-all consumption pattern since the chilled water system simply doesn’t run that much.  As a result, I will likely focus my attention on optimizing things like schedules, looking for throttled valves and dampers, and optimizing flow rates and set points.

Hopefully, this discussion has helped you understand that taking a look at energy consumption patterns via benchmarks and average daily consumption analysis, when combined with other indicators like operator interviews and field observations, can help focus your retrocommissioning and optimization efforts. In the next post, we’ll look at how this analysis also helped me understand why site energy consumption went up when common wisdom said it should have gone down when the electric heating equipment was replaced with gas fired equipment.

David Sellers
Senior Engineer – Facility Dynamics Engineering

This entry was posted in Uncategorized. Bookmark the permalink.

2 Responses to Understanding an Anomaly – Part 3 – A Hypothesis

  1. Lloyd Ricketts says:

    David, I work for a utility company as an analyst in the loss reduction division, I am interested in Normalizing monthly sales to reflect that of one calendar month rather than having it straddle into the next month. this would better enable us to gauge losses – (Net Generation minus Billed Sales) Generation is calculated at the end of each month therefore comparing it to billed sales which straddles two months is misleading. How do I approach this?

    • Hi Lloyd,

      This sounds similar to the issue we deal with when we get utility bills that straddle a month. For instance, the bill shows up on January 6th and reflects consumption from December 3 through January 2nd. The two issues are:

      1. (In my business) accounting will tend to call that the January bill even though most of it reflects December consumption.

      2. There are two two days of January consumption in the bill, the rest is December.

      Typically, I will want to compare consumption to information that is recorded on a true monthly basis, for instance, some heating or cooling degree day data is monthly, production records tend to be monthly, occupancy statistics tend to be monthly, etc.

      So, to solve that problem, I divide the billing consumption by the number of days in the bill, which gives me average consumption per day for the billing period. Then, I create the monthly consumption by multiplying the number of days in the month on each bill by the average daily consumption for the bill. For instance, in the example above, the December consumption would be calculated as two times the November average daily billing rate (December 1 – 2 are part of the November bill) plus 29 times the average December consumption based on the (December bill, which arrived in January).

      This is not perfect since it is based on average numbers, but is better than not doing it for the stuff that I do. The thing where Accounting calls the Bill that arrived in early January the “January bill” even though it mostly reflects December consumption can be particularly misleading if you don’t catch it.

      Anyway, a similar technique may work for your problem. If it helps, I wrote a paper about the technique I am discussing that you can down load for free from the California Commissioning Collaborative web site (www.CACx.org). Go to their library and search for Using Utility Bills and Average Daily Energy Consumption to Target Commissioning Efforts and Track Building Performance and you should find it.

      Bill Koran also developed a spreadsheet that does the normalization automatically if you give it the billing dates and consumption. It can be downloaded from the CACx web site too by picking the Resources for Providers drop down, then Tools and Templates, then Retrocommissioning. On the page that opens, pick Spreadsheet Tools and then look for the Utility Consumption Analysis Tool.

      Hope that helps,
      Thanks for visiting the blog, and sorry for the delayed response; I got kind of wrapped up in a report I needed to get out by the end of the year.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s