Energy Calendars

These radial calendars show the energy-use patterns of six Lake Flato projects in 2016. The hourly data we collect from our energy monitors can be plotted into a single graphic with a Grasshopper script and reviewed to understand energy patterns and anomalies that wouldn’t be apparent through numerical data alone.

How to Read the Calendars

Each dot represents an hour of energy, and each row represents a day, moving from midnight on the inside of the circle to 11 pm on the outside. A dot with a larger radius translates to higher energy use, typically ranging from about 0 to 50 watt hours per hour. The dots are colored based on the outdoor temperature at the time, from 40 degrees F in blue, to 100 degrees F in red. The calendars read clockwise starting with January 1 at the top.

Full Time Residences

The above calendar is apartment energy use. CPS provides some residences with smart meters that record energy use data on 15 minute intervals. A pattern of energy use at the same time each morning is visible, followed by a larger band of use in the evening. Daytime use is usually low, except on weekends. The color scale and heating loads reveal a milder winter in December 2016 compared to January 2016. Three energy gaps are visible in April, and two in July and December.

A small house in San Antonio, with a slightly larger area than my apartment and a leakier envelope. Air conditioning loads are most visible during occupied hours through the summer. Thermostat setbacks on weekdays keep daytime HVAC loads down. Heating loads peak on December 18, the coldest day of the year.

Part-Time Residences

In this residence in West Texas, it’s very clear when the house is occupied for a few days at a time. Temperatures are noticeably cooler than San Antonio, and energy use is extremely low when unoccupied except for a few small phantom loads.

This part-time house in central Texas, has visibly higher unoccupied loads and high HVAC loads, particularly during summer days and winter nights. The eMonitor stopped recording data for a few days in October.

Two Commercial Buildings

This medium-sized office headquarters in San Antonio reveals a typical commercial-use pattern; energy use picks up around 8 am, then slows down consistently around 6 pm, with a very visible 5-day work week occasionally broken by holidays such as Memorial Day and Thanksgiving. This eMonitor was installed mid-February; Lake Flato’s first eMonitor on a project of this scale. This graphic had to be re-scaled to be legible. The graphic at the same scale as the other residential projects in this post can be seen below:

The same office headquarters at the same scale as the others in this post.

Energy use from an open-air pavilion in north Texas. A solar array produces more energy than is used, making this pavilion a net-positive energy building. Energy use from water treatment peaked in mid-April during the heaviest rains of the year. Energy loads are very low for the remainder of the year.

Plotting energy generation from the solar array reveals a ring of solar energy produced during daylight hours. Days lengthen from about 7 hours of production in the winter to 11 hours mid-summer. The same rain event that caused high water pump use in April also limited solar energy production due to heavier cloud cover. Low cloud cover and long days through the summer make it the most productive time of year.

The Grasshopper Script

Each calendar starts with an Excel spreadsheet of hourly energy data and hourly temperature. If the energy monitor doesn't provide hourly outside temperature data, it can be downloaded from Weather Underground. The hourly data is then pulled from the file, and all blank data points are replaced with a default value. A circular array of points with 365 rows of 24 is created and flipped and rotated so the calendar reads clockwise with January at the top. Each dot is scaled down to an appropriate size and a color is assigned based on the hourly temperature. And of course, some radial and Gaussian blurring in Photoshop.