Link to an article on my Net-Zero Home Research

Everyone,

My research has finally been published in a peer-reviewed journal (Energy and Buildings). The process took six months total, not counting the time actually writing it. Unfortunately, if you want the full article, you may have to pay for it, depending on your access to Elsevier. 

http://www.sciencedirect.com/science/article/pii/S0378778813005446

Also, sorry it has been three weeks since Part 8 of my posts on my research. I have been very busy at my new job as a lecturer at UML, with weekends full of lesson prep and other work. Part 9 will come out, hopefully in the not too distance future.

Walter

Part 8 of my research on Net-zero Energy Homes in New England

As promised, today’s post will be all about how I modeled domestic hot water (DHW) energy consumption in the homes I studied for my PHD thesis. Please, stifle your yawns!

Before I dive into that, however, a couple of quick notifications.  I will soon have two articles published on this research. One is in the Sep/Oct issue of HomeEnergy Magazine (http://www.homeenergy.org/show/article/id/1903/nav/default). This one is all about the financial side of building and owning a net-zero or near net-zero energy home in New England. I know many of you are very interested in those results, so I strongly urge you click on over to HomeEnergy.org and take a look. Like most things in life, you will have to pay to read the full article. For copyright purposes, I won’t be able to simply cut and paste it into this blog.  Also, its going to be a while before I get to the financial section of my thesis anyway, so rather than wait several more weeks or months even, you can get the info right now at HomeEnergy Magazine.

The second article will be published in Energy and Buildings, a peer-reviewed international journal (http://www.journals.elsevier.com/energy-and-buildings/). The article has been approved and I am waiting for them to send me the final proof. Once I know it is published, I’ll let all of you know. That article is all about the energy results of my research project, with no mention of finances at all.  Okay, on to DHW modeling.

Remember, the whole point of creating my own custom energy model (CEM) was to have a model that could estimate energy consumption over a year in each of these houses, so that I could have something against which to compare the measured consumption. I couldn’t use “off the shelf” modeling software (see my previous post), so, using things like the ASHRAE Fundamentals Handbook and the DOE Building America Benchmark, I came up with my own models for the four categories into which energy consumption falls in a home: HVAC, DHW, lighting, and appliances & miscellaneous loads. Here’s an excerpt from my thesis with accompanying equations:

Energy consumed to heat a home’s DHW is dependent upon a number of variables. These include: the number of occupants in the home, the temperature of both the source cold water and the set point temperature of the DHW, the efficiency of the heating system, and significantly, whether the home uses a solar thermal system for DHW. The first step is to determine the daily DHW load in gallons, and for that a series of simple equations from NREL’s Building America Benchmark are used[1]. These calculate the average daily amount of DHW used for clothes washers, dishwashers, showers, baths and sinks. The sum of these load yield the total daily DHW load in gallons. Multiplying the daily load by the number of days in each month yields the monthly load.

In the equations below, “Nbr” refers to occupants (see the footnote on the bottom). Also, just click on the equations and a new tab will open up where you’ll actually be able to read them.

DHW1

 

Once I had the water consumption in gal/month, I converted it into liter/month using 3.785 l/gal.

Then, knowing the set point temperature of the hot water (which was one of my questions to each homeowner) and using the heat capacity (4187 J/kg-C) and density (1 kg/liter) of water, I determined how much energy is required to heat the amount of water consumed, i.e., the energy DHW load.

DHW2

Note that I assumed that tank and line losses amounted to 13% for each house. Thus, the actual energy required to heat the water would be 113% of the load. You’ll see I ended by converting the energy into kWh. BTW, these are screen shots from “MathCAD”, a very cool program (http://www.ptc.com/product/mathcad/).  I use MathCAD ver 15.0.  I tried using the latest iteration, which is called MathCAD PRIME 2.0, but I did not like the new interface.

Of the 20 houses I studied, nine had solar thermal systems, and most of those used them to heat DHW. So, I had to estimate how much of the required energy load could be met by these solar thermal systems, i.e., how much energy would the solar thermal systems produce and transfer to the hot water systems during the course of a year. I’m referring to the “solar fraction”–the fraction of DHW load met by the solar thermal system each month (you have to break it down by month; though I assume the homeowner consumes the same amount of water month to month, the amount of solar energy of course varies with the season…). More excerpts from my thesis follow:

The third step is calculating the solar fraction (SF) using the f-chart method.  The SF is the fraction of DHW load met by the solar thermal system each month.  This fraction greatly depends on the load, the size and efficiency of the solar thermal system, and the time of year.  Generally speaking, a correctly-sized solar thermal system may supply most of or even exceed the energy load during summer months, but produce only a small fraction of the load during the cold winter months.

The last sentence applies to homes in cool, snowy locations like New England. Homes in warmer climates (Florida, Arizona, Hawaii, etc.) may be able to generate all their energy for DHW from solar thermal systems year round.

            This step is broken down into two parts: part A calculates the average daily radiation on a surface of the solar thermal array for each month, Hc, in kWh/m^2; part B then uses the monthly values of Hc plus various parameters of the solar thermal system itself to calculate the SF for each month.  The equations in part A are essentially identical to those used in calculating Hc during the PV production modeling, taking care to use the tilt (β) and azimuth (aw) of the solar thermal system rather than the PV system (see equations 5 – 15 above).  The equations for part B follow, with “…PL representing the long term thermal loss per unit load and PS representing the long term insolation gain of the solar thermal collectors” (Goswami, Kreith, & Kreider, 2000).

I’m not going to show the work to calculate Hc, its simply too long. However, part B may prove interesting to some people, so here are the equations:
DHW3
Note that in the equations above, I’m using tech specifications (FrUc, etc.) for the solar thermal system in House W.

The fourth and final step calculates the DHW load, taking solar fraction and energy factor of the hot water heater into account. Ld (monthly DHW load) is multiplied by (1 – solar fraction) to get the DHW load required to be met by the home’s hot water heater.  For homes w/o solar thermal, the SF = 0 for every month.  Finally, these values are divided by the hot water heater’s Energy Factor (EF) to convert from load to consumption, and by 3.6*106 to convert into kWh:

DHW4

As an example, the figure below gives the results of these DHW calculations for House W, which has a solar thermal system.  Note that between March and October the DHW load goes “negative” in this graph. This means that the solar thermal system can supply a greater than required amount of energy for DHW during this period, and hence the electric back-up system should not have to supply any energy during the same time period.

DHW6

 

That’s how I modeled DHW energy consumption.  I hope I did not make this more complicated than it needed to be for you. Next time, we’ll talk lighting load, which will be a little less involved, promise.
Walter


[1] Note: the author has modified these equations by substituting the number of occupants of each house for the number of bedrooms. The primary reason is that in several of the homes studied, the relationship between number of bedrooms and occupants varies greatly. For example, House A has seven occupants but only three bedrooms, whereas House N has five bedrooms but only two occupants. Clearly, the people in House A would use more DHW than those in House N, but if number of bedrooms is used as the variable, then the equations would yield the opposite results. In the cases where number of bedrooms equals number of occupants, then substituting the variables makes no difference to the results.

 

 

Part 7 of my Research on Net-zero Energy Homes in New England

Hello again.  Last time, I talked about how I modeled PV production for the PV systems installed on the homes I studied. Today, I will introduce the discussion on how I modeled energy consumption for the homes.  To save some time, I’m simply going to use some verbiage right out of my thesis (in italics).

Keep in mind that my research analyzed not just the energy performance of the homes I studied but also their economic performance, as well as the carbon dioxide emissions ramifications of these homes. I know, based on feedback I’ve received, that many readers are more concerned with these other topics than they are about the actual details on the modeling. Therefore, though I am going to give it substantial coverage in this blog-version of my research, I will try to keep things moving so that we can get to the other topics before the next Olympics…

I used the following paragraphs from my thesis to introduce the methodology I followed for energy consumption modeling:

The author created a “Custom Energy Model” (CEM) that uses as input basic building and system specifications, geographic location, and occupant behavior, and that yields estimates for a home’s monthly and annual energy consumption.  These include energy consumed for domestic hot water (Edhw), lighting (Elight), appliances and miscellaneous electrical load (MEL) from cellphone chargers and modems (Eappl), and heating, ventilating and cooling (Ehvac).  

Essentially, all the energy consumed in a home can be divided into these four categories.  The very first time I used this method was when I was working on UMASS Lowell’s 2011 Solar Decathlon entry…see http://www.4dhome.us, though that effort was far less sophisticated than what I ultimately used in this research.

The initial homeowner survey contained about 50 questions that provided the necessary information to “run” the CEM.  The equations in the following subsections are used in the CEM and generally follow methods outlined in the ASHRAE Fundamentals Handbook (ASHRAE, 2009), the DOE Building America Benchmark (Hendron C. E., 2010), and in Goswami (Goswami, Kreith, & Kreider, 2000).  Note that these equations calculate energy loads for each hour of each day of the year (8760 values). Hourly calculations were undertaken in order to analyze the specific performance of various systems. They were summed over time to yield monthly and annual loads.

The above paragraph shows that I did not reinvent the wheel to come up with my energy consumption predictions, but rather simply applied existing principles, assumptions and techniques.

There were two reasons for developing an original energy model: necessity and reasonableness. It was necessary, as the more complex energy models currently available (Energy+, Energy10, EnergyGauge, etc.) require very in-depth specifications.  However, the author did not have access to that kind of data for these homes, and typically, designers and homeowners did not either.  Hence, a model was required that would work with only limited information available on each house.  

Early into my PhD work, I had contemplated using an “off the shelf” model (there are literally dozens out there) to provide predictions of energy consumption against which I could compare measured consumption.  However, as I stated in my thesis, I soon learned that I really did not have sufficient details on the homes’ specifications which all of these programs required.  I thought about asking the homeowners for more details, but decided against that, in fear that several would decide this “volunteer” involvement in research was becoming too much of a hassle and drop out.  I did not want anyone to suffer “volunteer fatigue”.

It was also a reasonable approach because occupant behavior and weather variation have an overwhelming impact on a home’s energy consumption.  These completely mask the relatively minor deviations caused by using more detailed specifications in a model.  Hence, since the purpose of this research was to measure energy performance relative to broad goals, a simpler model was more appropriate.

I was looking for big picture results, more concerned about accuracy than precision. If a home was designed to be net-zero, did it achieve net zero? If my model predicted a near net-zero home would consume 20,000 kWh in a year, did its actual energy consumption come close? If not, what were some reasons? Please keep in mind: just like with my own PV production models (see my previous post), I was not trying to prove my models were better than existing models.

The author used the CEM to produce an estimate of annual consumption for each house, and then compared this estimate to the one used in designing the home, if available.  Finally, he compared these two estimates to the house’s actual consumption over the 12-month monitoring period.  

I did ask each homeowner to provide the original designer’s energy consumption prediction for their home, though less than half were able to provide it to me.  In subsequent posts that include energy consumption results, you’ll see the “homeowner provided” predictions for those houses displayed along with my model’s predictions and the measured consumption.

Ok, that’s the intro on energy modeling. I’ll take on the first category of load–DHW–in my next post.

Walter

 

Part 6 of my research on Net-Zero and Near Net-Zero Energy Homes

As a quick reminder to everyone–this series of posts is my attempt to share the research I performed to earn my PHD in energy engineering. I studied 20 homes in New England–10 designed to be net-zero energy, 9 designed as “near” net-zero energy, and one control house (an Energy Star certified home with a HERS of 67). I collected a variety of construction specifications and occupancy statistics on each home, and collected energy production, consumption and cost for each house for 12 consecutive months.  The homeowners were all volunteers, with their names and exact addresses withheld for privacy.  Please see my previous postings to learn more about the objectives of the research, my data collection methodology, and details on the homes themselves.

This week, I am going to briefly dive into how I modeled the PV production of the systems installed on the homes.  First off, 18 out of 20 of the homes had PV systems. They ranged in size from 1.1 kW to 14.4 kW.  Table 1 below lists some of the details for each system.

Table 1

House ID

On-site Electrical System

Electrical System Power
(kW DC)

PV Module Model

Solar Thermal System

A

PV

3.6

CS6P

B

PV

3.6

CS6P

C

PV

3.6

CS6P

G

PV

1.1

Evergreen 190

H

PV

7.6

ES-200

Yes

J

PV

4.8

EC-115

Yes

K

PV

12.4

SunPower 230

L

PV

3.7

UNK

M

None

N/A

N/A

Yes

N[1]

PV

12.1

SunPower 238

Yes

O

PV

10.1

SunPower 225

Yes

P

PV

6.9

CS6P-230P

Q[2]

PV & Wind

4.0 PV;
10.0 Wind

UNK

R

None

N/A

N/A

S

PV

14.0

SPR-225-BLK-U

T

PV

4.2

Es-A-210-fa3

Yes

V

PV

6.9

CS6P-230P

W

PV

7.6

ET Solar 230

Yes

X

PV

3.7

UNK

Yes

Z

PV

7.1

UNK

Yes


[1] House N had four, independent PV systems in four arrays. Two faced south, and one each east and west.

[2] House Q has a Bergey 10K wind turbine. It is also the only home to have a dual-axis tracking PV array. All other homes used fixed, roof-mounted arrays.

I could have just used DOE’s PVWatts on-line calculator to predict annual energy production for each of these systems.  However, for a number of reasons, I decided to model their output myself, and use PVWatts simply to verify that my results were at least similar.  To be clear: I had no intention of proving my model produced more accurate results than PVWatts. To the contrary, since PVWatts is recognized in the industry as being an easy and accurate method for predicting annual (and monthly) output of a PV system, I thought it wise to use this tool as a means to check that I had not made some error along the way.

Modeling PV energy production (EG) is relatively straightforward and highly dependent upon three factors: solar angles, system specifications, and shading.  The angles include the latitude of the house and the azimuth and tilt of its PV array. Homeowners provided the tilt and azimuth angles while I determined the latitude for each house by inserting its street address into Google Earth.

For each house, I calculated the monthly average irradiation per day, HC, in kWh/m2-day, using these angles and the irradiance data from the TMY3 data set in a series of equations. (See my complete thesis for these equations, beginning on page 56).  They include equations to calculate declination angle, clearness index, cosine of the zenith angle, and several other rather esoteric parameters.

I used software called MathCAD for these calculations. The graph below is simply a cut and paste from the MathCAD file for House A’s PV system.

Image

You can see that in month 12 (December), monthly average radiation is at its lowest, and that it peaks in July. No surprises there.

HC, multiplied by the module area, the number of modules in the array, module efficiency, and the number of days in each month, yields the energy produced per month in kWh:

Array output per month, DCm, kWh:

Image

η = module efficiency

Area = module area in m2

N = # of modules in array

mdm = # of days per month

Then, this DC energy is converted to AC output by multiplying by the “derate” factors, such as inverter efficiency, module mismatch, etc.. I used the same categories of derate factors as PVWatts, using specifications for each system.

Finally, I multiplied each month’s AC output by the corresponding monthly shading factors, which I gathered for most of the systems during site visits using a Solmetric Suneye device. (Just a quick explanation for those who don’t know. If a PV system is completely unshaded during a month, that means no shadows from trees or other obstructions fall on the PV modules during any time of day. The SunEye estimates how much shading would occur each month, and produces a percentage of system output based on that shading.)

Here is what the monthly output for House A looked like:

Image

Below is a different graph of the same data (the blue bars), but arranged from Sep 2011 through Aug 2012 (the period during which I collected actual production data). You will also see the PVWatts output and actual production from House A’s system:

Image

In case the graph above raises some questions in your mind…House A’s system had a malfunction in Jan 2012, hence its measured production was less than half of either prediction. April 2012 was an exceptionally sunny April compared to “typical” Aprils. Hence, the measured production for April was nearly 40% greater than either model’s prediction.  Such a large difference between measured and predicted was also apparent in most of the other houses, since New England is a relatively small geographic area and everyone was experiencing very sunny weather that particular April.

For most systems, my model and PVWatts agreed, and generally matched the measured production. I will talk more about the results in later posts.

Well, that’s it really for the modeling of PV output. I did have to develop a separate model for the one double-axis tracking system I studied, which was really easier than it would seem (fewer angles involved since the tracker keeps the sun’s rays perpendicular to the module surface at all times).

Next post, I will begin to discuss how I modeled energy consumption. THAT is a much longer process, and I devote a considerable portion of my thesis to creating the models–37 pages! FYI, I’ll be travelling this coming week, so I may not get to the next post until after the 18th.  If you want to read ahead, you can check out my thesis on-line. Its full title is Energy Performance of Net-Zero and Near Net-Zero Energy Homes in New England.

My Research on Net-zero and Near Net-Zero homes in New England–Part 5

Last post, I described some of the details on the 20 homes that were included in my research project on net-zero and near net-zero energy homes in New England.  Today, I’ll describe some of the common features of these homes, and also explain how I collected the data I used in my research. I’ll try not to get too bogged down in details here, but forgive me if I do, please!

Insulation levels: Most homeowners understand that the more insulation they have in their home, the lower their energy bills ought to be.  Most localities have building or energy codes that specify minimum levels of insulation for walls, roofs, floors, etc.  Typical new homes are built to code, and hence meet these minimums.  However, homes designed to minimize HVAC energy loads will have high amounts of insulation in their walls or “envelope”.  This of course includes homes designed to be net-zero or near net-zero. Table 1 below shows the average wall and roof R-values (in F-ft^2-hour/Btu) for the homes in my research project.

Table 1

House ID R-Value Walls R-Value Roof
A 21 34
B 21 34
C 21 34
G 40 60
H 26 38
J 24 38
K 34 57
L 21 50
M 31 45
N 30 45
O 31 50
P 68 75
Q 45 60
S 60 72
T 27 52
V 44 75
W 45 60
X 43 50
Z 40 60
Average 35.4 52.1
R (control) 19 30

Tight envelope and air “sealing”:   Buildings–homes included–typically “leak” air between the interior and exterior. If they have a lot of leaks, they’re labeled “drafty”.   For instance, if you’ve ever sat next to a single pane window in an old colonial or Victorian style house on a New Hampshire winter day, you’ve probably felt a draft coming from that window.  Newer, better-built homes won’t feel drafty, but they still leak.  For example, we live in an Energy Star certified home with above-code insulation.  However, on a cold windy day, I can still feel cool air coming into my house through some electrical outlets mounted in exterior walls.  The wind is pushing air through small gaps in the exterior sheathing, and that air works it’s way through the various layers of the wall until is finds its way into the interior, via the slots in the outlets.

One metric of how well a home is sealed against air “infiltration” is called air changes per hour (ACH).  An ACH of 1.0 means that there is enough air moving to replace the entire volume of air in the house in just one hour.  That would not be good from an energy consumption perspective, since the heating system would have to keep warming up the cooler air as it moved into the house. Of course, you need some fresh air entering the home otherwise the air quality would deteriorate.  So, while a low ACH will reduce load on a home’s HVAC systems, you don’t want it to be zero.

There is a tremendous amount of variability in ACH rates across the housing stock in the US, depending on age, location, style, etc., etc.  However, having a well-sealed envelope is a basic design parameter in a net-zero home, and, based on the ACH rates listed below, all the homes in my research were very well-sealed indeed:

Table 2

House ID Natural ACH (heating)
A 0.07
B 0.1
C 0.1
G 0.1
H 0.1
J 0.092
K 0.061
L 0.1
M 0.04
N 0.21
O 0.1
P 0.04
Q 0.09
S 0.05
T 0.1
V 0.05
W 0.1
X 0.1
Z 0.07
Average 0.085
R (control) 0.14

Windows: Like high amounts of insulation and excellent air-tightness, having high-quality windows is nearly a prerequisite for a net-zero energy home.  A lot of energy can be lost through windows–not just through leaky frames but passing through the window pane and frame themselves via conduction and convection. The U-value of a window is used for comparing the heat transfer parameter of a window (U-value = 1/R-Value). The lower the U-value, the less heat will be lost through that window. I can hear you now–“I bet the homes in Walter’s project had really low U-values, too”. Well, you’re right! Here they are:

Table 3

House ID Average Window
U-Value (Btu/ºF-ft2-hour)
A 0.3
B 0.3
C 0.3
G 0.29
H 0.3
J 0.33
K 0.17
L 0.25
M 0.22
N 0.25
O 0.27
P 0.22
Q 0.17
S 0.25
T 0.29
V 0.24
W 0.21
X NA
Z 0.12
Average 0.24
R (control) 0.33

Keep in mind these were average U-Values. Every home used multiple styles of windows, and each would have had a different U-value.

Lighting: All but one of the houses used at least 85% LED or CFL lighting.  This is pretty straightforward, in that LEDs and CFLs typically use 70-80% less energy than a comparable incandescent light, hence they are a natural feature for any net-zero or near net-zero designed home.

Appliances: All the homes used Energy Star certified appliances.  Many used very efficient appliances, electric or gas.

Passive solar design: All the homes’ long axes ran east-west, meaning they had long south-facing walls.  In New England, where heating load is far greater than cooling, long south walls mean more solar gain in the winters, which will reduce heating load.  This heat gain comes mostly through windows in the south walls.  The homes had interior shading devices (curtains, blinds, etc.) to cut the solar gain during the summer, to lower cooling load when necessary. Many of them were designed to maximize natural or “day” lighting, too, to further reduce lighting energy load.

I think you get the idea.  All of these homes were designed to minimize energy consumption, and hence they all had high amounts of insulation, were very well-sealed, used excellent windows, CFL and/or LED lights, and highly efficient appliances, and were built with their long axis running east-west to maximize solar gain. These are all features common to net-zero homes, and it is no surprise they were included in these homes built in New England.

Okay, just a little bit on my data collection methodology.  Once a homeowner signed on to participate,  I sent him or her an “initial survey” to collect basic construction information, occupant statistics, cost information, HVAC systems make and model number, etc.  This came to about 50 questions.  Some of the owners sent the completed survey back within a week or two, others took months and months.  Beginning with September 2011, I collected energy consumption and production measurements in one of three ways: homeowners sent the information directly via email; homeowners sent a copy of their actual utility bills, from which information was collected; or the information was read off production/consumption reporting websites to which homeowners had granted me access (e.g., Solectria Renewables’ “Solren View” website). Depending on availability and applicability, data included: billed electrical and natural gas usage; monthly cost for electricity and natural gas; quantity and cost of any other fuel purchased or consumed that month (e.g., gallons of propane); and energy production by renewable generation systems. Eventually I collected 12 months of energy data on 17 out of the 20 homes, with three homes providing fewer than 12 months because they started a few months late. I also gathered energy cost information from 16 out of the 20 homeowners.

The surveys and monthly data gathering was just the tip of the data iceberg, so to speak.  I also placed temperature and humidity data loggers inside and outside 13 of the homes, to record those parameters during the project.  I installed eMonitors in two of the homes to monitor circuit-level energy consumption (had I had any funding, I would have installed these great devices in all the homes). Two other homeowners installed eMonitors on their own and provided me with their data.  Then I went to most of the homes and, using a Solmetric SunEye device, measured the loss due to shading of their PV systems. I also downloaded lots of weather data from the NWS and other providers to use in my modeling.  Finally, I downloaded Typical Meteorological Year (TMY) data sets from NREL’s site to use in my PV production models.  Lots and lots of data.

That’s clearly enough for now.  Next post, I will get into my energy modeling, starting with PV production.  After that, I’ll get into the consumption modeling, and then start presenting some results.  Looking further ahead, there’s the economic results to discuss, as well as the carbon dioxide emission simulations I performed, then my conclusions and recommendations.  So, lots more to come people. I hope I am keeping it interesting for you!

Walter

 

DOE Overestimating Impact of Energy Efficiency Standards on Appliance Prices | Articles | Distributed Energy

A little off-topic for this blog, but I wanted to share because I think energy efficiency is the cornerstone on which our energy future will be built.

This is a good news story.  Despite the good people at DOE not quite getting their estimates right, their energy efficiency standards for appliances have not only saved consumers money by reducing their utility bills, but have NOT driven up the cost of those appliances (in fact, they’ve dropped over time).  So, thanks to the DOE’s standards, we are using less energy (good for everyone) AND saving money (good for each individual).

A true win-win. Kudos to DOE and their Energy Star program!

Walter

 

DOE Overestimating Impact of Energy Efficiency Standards on Appliance Prices | Articles | Distributed Energy.

Lowell power curtailed despite high power demand during recent New England heat wave

Very in-depth article on why power from a wind turbine farm in northern Vermont had to be curtailed (ie, its output deliberately reduced) to ensure stability of the local grid during a recent heat wave. Transmission line inadequacies and not having all the needed interconnection equipment were to two causes.  FYI, it starts out from the perspective of Vermonters who wonder why these turbines were not used but diesel-powered generators were, then ends with a clear explanation from ISO NE.

I recommend reading this piece if you are involved in utility-scale renewable energy projects anywhere.

Walter

PS–smaller renewable energy projects (like PV panels on homes or businesses) on their own do not usually cause these types of grid stability problems.  Another reason to push for more distributed energy in our future electrical system.

Lowell power curtailed despite high power demand during recent New England heat wave.