There are many reviews of the Nest Learning Thermostat but they have primarily focused on installation, features, and user experience. Putting all features of Nest aside if the Nest does not live up to the promise of saving energy none of its features or “gadget-y coolness” matter in the end. To truly review the Nest I have analyzed my energy usage during the first three months of having the Nest installed. This is after the Nest had enough data to activate all of its energy saving features which include:
- Time to Temp
Let’s start with the raw data. The data represents May, June, July, and August of 2011 and 2012. The chart below shows kWh for each month and the average high temperature for the each month comparing 2011 to 2012.
Total energy usage year to year comparison
As you can see there is a significant drop in kWh from 2011 to 2012, but there are several factors that need to controlled for. In order to evaluate the energy savings of the Nest I will make the following assumptions:
- There has been no significant change in my lifestyle year over year for the three month period in questionα
- For every one degree increase in the average temperature it increases my energy usage by four percentβ
In order to normalize the data with the above assumptions I took my 2012 usage and adjusted it to 2011 temperatures. For example in July 2011 the average temperature was 5 degrees warmer than July 2012, hence I increased the kWh in July 2012 by 20% to account for the temperature differential. The chart below shows kWh usage normalized to 2011 temperatures for all four months.
kWh normalized to 2011 temperatures at four percent per degree
So even after controlling for temperature differences year over year there is still a significant decrease in energy usage. The chart below shows the percent change in kWh for each month year to year after normalizing the data.
Percent change in kWh for each month year over year
While this isn’t a perfectly controlled analysis, given the percent change in energy usage, I think its safe to conclude that the Nest thermostat offers energy savings given my conditions. Over the four month period this equated to a savings of $305 based on the normalized numbers which means the Nest paid for itself in four months. Its worth noting that I normalized the data across all kWh not just those kWh used for air conditioning hence these savings are a conservative estimateγ.
Opportunities for Improvement.
Overall I am very happy with the Nest thermostat and I believe it lives up to the energy savings promise based on this analysis, though I am hesitant to give those savings a hard number based on the shortcomings of this analysis δ.
I could probably extract even more savings from the Nest if it provided geofencing capabilities and could detect if I left the house or was approaching the house based on the location of my iPhone. When I purchased the Nest I thought “The Leaf” feature, which shows you a green leaf if you set the temperature to one that is energy efficient, would alter my behavior, but ultimately I just ignore it and change to the temperature that is most comfortable.
It will be interesting to see what reporting Nest will offer once it has year over year usage data for a large number of users. I think one of the key factors that makes it easier for me to yield energy savings from the Nest is the fact that I have a single person household. It is concerning to me and seems like an obvious miss that Nest does not collect more demographic information when users register their thermostat. Demographic data such as number of persons in the household, home square footage, etc… could really enhance the reporting they can provide to users as well as improve their algorithms. The killer feature for the Nest that could really deliver on its value proposition would be the ability for Nest to pull my energy usage data directly from my provider analogous to how Mint can pull bank account information.
I was able to easily recoup the cost of the Nest in four months, but in my opinion there is enough value in its utility alone to justify purchasing a Nest Thermostat.
This post was updated to include four months of data. When originally published it included only three months of data.
Several people have asked if I was using a programable thermostat before installing the Nest. I was using a Honeywell Chronotherm IV Plus, which is a programable thermostat, but I never programmed it properly. So it should be noted that a significant part of any actual energy savings probably comes just from making use of a programable thermostat as intended. Though I would also suggest that a key part of the Nest’s value is in the fact that most people never program their thermostats even if the feature is available, hence the EPA dropping off the Energy Star rating from programable thermostats.
To respond to several other questions I have received and that are worth factoring into my results:
I have a single story home with one Nest unit that is ~1700 square feet. My house was also built in 1922 which means its not very energy efficient with lots of opportunity for air to leak in out of the house.
That is to say my non air conditioning electricity usage such as wash/dryer, PS3, TV viewing hours, dishwasher, etc…are given as equal year over year.↩
This is a strong assumption, I found this site that mentions that for every one degree you increase your thermostat it increases your energy costs by 4%. I actually need the inverse of this for my analysis, but I was unable to find any information on what a 1 degree increase in outside temperature does to energy costs. I also assume this is very dependent on where you have the thermostat set relative to the outside temperature. Furthermore, I would also assume that the relationship can’t be linear, for example a 1-3 degree increase in outside temperature causes a 4-5% increase in energy usage, but a 4-6 degree increase in outside temperature causes a 7-8% increase in energy usage.↩
It is assumed that higher outside temperature do not significantly impact usage of other electrical appliances. Air conditioning accounted for 49%, 52%, and 54% of my total electricity usage for May 2012, June 2012, and July 2012 respectively.↩
If you have any thoughts on how I could improve this analysis please let me know.↩