ESP167FinalPaper

By: Leslie Canones, Kara Chiu, Steven Dunbar, Arnold Ki, Clinton Lam-Song, Brian

Shaw

Table of Contents:1

-Abstract

-Introduction

-Problem statement

-Background

-Methodology of project

-Results/Deliverables

-Conclusions

-Recommendations for future action -Appendices

Abstract:

Energy Management in buildings is an important topic to be addressed because managing

and reducing energy consumption can save money and help reduce greenhouse gas emissions.

Therefore, understanding what the most prominent factors of energy use are can lead to more

meaningful policy change to reduce consumption.

2

This project will look at the energy used in Vet Med 3B and compare that data to

estimate usage in three other buildings on the UC Davis campus. The first key point is to break

down the data given on Vet Med 3B into categories of use, such as plug load, HVAC, lab

storage, and lighting. The second key point is to obtain a method of comparison between Vet

Med 3B, which has multiple energy meters for data breakdown by type, to the other buildings,

which only have a single energy meter for the entire building. For this, considerations of

building type and breakdown of lab versus office space will be accounted for, as well as HVAC

usage, which is not included in the meters (because it is done at the central plant). From this data

collection our group will analyze and make suggestions for future building energy use.

Introduction:

Over the history of human civilizations, the building design and its efficiency have

evolved. The main purposes of buildings are still to provide hospitality, amenity, and a working

environment. However, as humanity entered into the 20th century, the use of electricity became

prevalent throughout all buildings. This energy use became known to consume a lot of energy

and emit great amounts of carbon dioxide, which has become one of the most abundant

greenhouse gas today. The general components contributing to building energy consumption are

lighting, plug loads, HVAC (heating, ventilation, and air conditioning), and the heating and

cooling of water. Compared to other buildings that are composed of office and classroom spaces,

laboratory buildings consume much more energy because of dominant HVAC usage to ventilate

the lab space.

Problem statement:

3

Our problem statement consists of three things. First, from the data that was given to our

team by Mr. Starr of the Vet Med 3B building we were to compile the building’s data and make

it comprehensive and comparable to the other buildings. Which in our second objective, our team

was to compare and extrapolate the energy uses in other buildings such as the Ghausi

Engineering Hall, Chemistry Annex, and Earth/Physical Sciences buildings. Finally, we were to

create recommendations for energy savings, based off these comparisons and analyze their

feasibility.

Background:

Buildings account for the largest portion of energy use in the United States. Within these

buildings some of the contributors of energy use can be found in lighting, space heating and

cooling, computer/electronics and other plug loads such as fridges and lab equipment. Globally,

buildings account for about ⅓ of primary energy use. Furthermore, in the United States,

buildings account for approximately 72 percent of total electricity use and 36 percent of natural

gas use. In addition, U.S. building sectors contributes 10 percent to GHG emissions via fuel use,

contributing about 9 percent of the world’s carbon dioxide emission. However, amongst the type

of buildings, laboratory buildings consume large quantities of energy- often 3 to 4 times more

than offices or classroom per square meter. This is because of the different regulation standards

on lab buildings compared to office and classroom spaces. Furthermore, employee behavior for

all buildings contributes to energy use. However, the Vet Med 3B uses sustainable strategies to

decrease energy load by maximizing natural lighting, using ambient free cooling to decrease

freezer farm energy demand, and by having individual meters to keep track of different types of

energy use within the building.

4

Methodology of project:

Our project started with Mr. Starr giving the team a tour of the building so that we could

visualize how the building was set up as well as inform us about the different energy uses of the

lab building. After that, Mr. Starr provided our team with the data from the Vet Med 3B

building. Our first course of action was to review the data generally and then to simply plot each

row using computer programs such as Matlab and Excel. However, this presented a few

problems. In particular, some rows in the data were misaligned or missing entirely. Since the

data is far easier to analyze if it is in consistent time intervals, this had to be fixed first. Our team

was able to accomplish fixing this mistake by using both visual and programming checks for

empty cells. But this proved to be ineffective because of the row numbers being skipped.

Eventually, our team checked for empty cells that were not part of fully empty rows. These

errors were corrected either by combining rows of data points that were either 1-second apart

5

from each other, or repeating previous data points for entirely missing times. The corrections

help make the averages be more consistent.

Once all of these problems were corrected, we separated the data into averages for each

time point between all weekdays and weekends (see Matlab appendix). We then added various

meters together as indicated by Mr. Starr (see appendix). For example, to get total metered

lighting, our team would add all the lighting meters in the building together to get total lighting

energy use. Additionally, once the last data points were separated into weekdays and weekends

and matched up, this allowed a quick check on the integrity of the data. This was to double check

that the data would not have any outliers or data points that did not match up. After the data was

acquired and verified, both Excel and Matlab made it easy to graph.

The graphs were reworked various times to make them easier to read and interpret, and to

correct errors that were a result of the data. In particular, various colors were introduced on all

graphs. Also, on each graph, the lighting percentage tends to go below zero in a few places

because of different data points taken by each meter; the two main meters take 15-minute

averages of power, while the sub-meters take instantaneous power. Because the lighting is all of

the sub-meters subtracted by the main switchboard meter, it is the most noticeably affected. As a

result, the lighting use data tends to spike dramatically at the beginning and end of the day, when

there is the most rapid change in usage. All of these averages were then summed and

manipulated to get the percentage of what each use contributed to the total.

Comparing to other buildings was where the most significant challenges were faced.

Although total electrical data was provided for each building, this electrical use is not the entire

story: the steam and chilled water use from the central plant also needed to be included in the

totals. Unfortunately, the Vet Med 3B did not have currently operating meters for these uses. As

6

a result, we needed to compare estimated data, which when added to the electrical data indicated

that the HVAC use could be as high as 92% of the total building load. In addition, we were also

informed that there were other loads on the HVAC use lines, such as an autoclave and an

industrial washer. This means that although they do not contribute to the HVAC energy use,

their usages are added into the total HVAC use line. However, even if these other usages are

taken out of the total HVAC, HVAC would still account for more than 50 percent of building

energy use.

*Still working on the comparison. I’m going for kwH/sF/day in April average because obviously I can’t average the yearly use.

7

Results/Deliverables:

8

9

Recommendations:

For all recommendations, initiatives are analyzed building by building as noted; because

the Vet Med 3B building is quite new and efficient, some programs would be more effective for

older buildings, but do nothing for Vet Med energy use. These recommendations have the

potential to help future prospects of new lab buildings and other buildings looking to invest in

energy efficiency. There are three types of energy use reduction recommendations: policy,

operational, and technological.

Policy Recommendations:

As a general rule, policy recommendations are the cheapest to implement, and for Vet

Med 3B represent the largest potential savings at this point. Even if programming/personnel

time costs for the last recommendations are low, they could have short payback times depending

on behavior changes, and could increase knowledge about the subject of energy use, leading to

savings in other areas on campus.

Our first recommendation is to get labs to reduce receptacle use to a minimum: this

includes reorganizing refrigerators, turning off unneeded equipment, and removing other items

that take standby power. This is applicable to all buildings generally, but particularly to

biological lab buildings such as EPS and Vet Med because we expect a higher use of freezer

farms for biological matter. The graph of use for Vet Med 3B reveals that freezer farm and lab

receptacles make up 26% of the building use, giving at least the potential for significant savings.

This policy could be limited by contamination factors, how well the use is already optimized,

and by ease of use constraints (no one is going to walk across the building multiple times per day

for freezer use). An indicator of its effectiveness might be a measure of freezers per room/area: a

higher number could indicate overuse, and thus available savings. If recommendations are

10

applied, this can result in greater energy efficiency with refrigerators and freezers in close

proximity and unnecessary machines turned off.

The second policy proposal our team had, is to make sure to turn off lights and other

equipment on weekdays, when they are not needed. It can be seen from the weekday/weekend

comparison that the lighting, HVAC, and total weekend curves are significantly lower than

weekday night use. This shows that lighting, HVAC, or other equipment is not being turned off

on weekdays as compared to weekends, in the nighttime. Although the HVAC reduction in Vet

Med 3B is limited by wing-to-wing reductions (discussed later), this does not account for the

entire difference in the total. In addition, it is likely that this is simply from workers working

later, but even in this case, the weekday total never equalizes with weekend use. These savings

for Vet Med 3B are significantly smaller because they are only targeting particular times that

already have low use, but this could be a larger factor of use in other buildings that have less

automatic control.

The first policy recommendation was to eliminate unnecessary devices that are not used

by checking if appliances are up to date and address how much energy each device generates. If

machines are necessary for use but not used often, then they can be turned off when unused and

keep turned on only when it is in use. Second, buildings that want to save energy can purchase

more energy efficient devices that use less energy by purchasing more energy efficient devices

such as energy star equipment. To convince building managers to install energy efficient

devices, they can show the cost benefits from energy savings to offset the initial cost of the new

energy efficient devices. Another option for policy initiatives is to retrain and inform staff about

energy consumption by encouraging them to turn off devices such as computers and lights at

night and during the weekend. There could be a group leader or manager that checks whether or

11

not people follow the suggestions to reducing the building’s energy use. This can be an

incentive to follow through with turning off devices and lights so that workers are not singled out

by not following energy saving methods. This can become applicable in the staffs’ daily lives

outside of the office as well. If worker’s mindsets lean towards energy saving, it can help spread

a similar mindset to others.

Lastly, a cross-category initiative would be to provide direct feedback on the use of

energy, similar to the campus energy feedback system currently being implemented.

Piggybacking on that system’s implementation may allow lower cost, and could provide

notifications and reminders to users as they leave. This would naturally have a short payback

time, especially if implemented elsewhere first, but its benefit to total energy reduction may be

minimal, as discussed in the first recommendation.

Operational Recommendations:

Our first operation recommendation would be to turn down HVAC systems on a room-

by-room rather than wing-by-wing basis. In all, the HVAC can only be turned down 33% from

its nominal capacity, but given that HVAC is the majority of the power use, this would be

considerable if possible. Unfortunately, there are many technical hurdles to this

recommendation, including room-by-room adjustable ventilation and reliable occupancy sensors.

If these could be done without dramatic reconstruction, this could reduce the dramatic tail-off

time of weekday HVAC use, and again recognizing that total HVAC use accounting for steam

and chilled water is even higher than graphed, could have a decent reduction in use. This benefit

would be much higher in other buildings that are currently less automated or do not turn down at

all. However, these buildings would likely require more expensive renovations to accomplish

this feat, so this recommendation would have to be analyzed on a building-by-building basis. A

12

cost-benefit analysis will most likely be necessary to find out what would be the best move for

each building.

Our second operational recommendation is to utilize occupancy sensor controlled plug

strips. These would automatically shut off the power when no use is detected, reducing vampire

power and unnecessary light loss. Because Vet Med 3B already has ambient light sensors, this

would be largely redundant, but it would reduce overnight lighting use. However, in other

buildings without sensors, higher lighting use, and other vampire plug loads, this may have a

shorter payback time. The sensors would be able to turn off when there are no workers in the

room, but to make sure the sensors are working properly it would be highly recommended that

maintenance regularly check on the sensors to see if they work at all in turning off the lights

when there are no people, and to see how efficiently they work. This will be helpful to give

insight to other buildings if the Vet Med building installed these sensors so that buildings can

decide whether the sensors work well enough to invest in them. There is a possibility that these

sensors are not cost effective if they do not help with the reduction of energy use and costs. It

will be the maintenance or management’s responsibility to record the usefulness of the sensors.

Our last recommendation is again partly operational, partly technical. Reducing

additional exhaust losses by closing fume hoods when not in use would lower the amount of air

that is cycled above what is required by code. This naturally requires fume hoods to be

applicable, but it could be cheap and effective. It could be as simple as fume doors that slowly

close on their own. They could have a lock setting if necessary, but the recommended setting

would naturally be to close. One of the main aspects that would make it not useful is that many

of the gaseous chemicals that are filtered out of the lab by the fume hoods could potentially be

stuck in the lab if the fume hoods do not completely take out contaminated air. It could

13

potentially be a dangerous drawback if the fume doors are closed before all air is taken out.

Due to Vet Med 3B’s highly efficient HVAC system, the most substantial operational

recommendations only apply to other, older buildings. Improving HVAC systems in older

buildings leads to the most returns in efficiency in terms of electrical use reductions.

Specifically, improving and renovating building insulation materials to reduce temperature

management needs may provide a large boost in energy efficiency of lab buildings, up to the

ventilation limit, if other leaks are unknowingly releasing chilled air in addition to the imposed

limit. This would require specific building audits, but given the exceedingly high use of HVAC

systems, could be cost effective. Older buildings are able to look to Vet Med 3B for inspiration

or ideas to use for reducing their HVAC system’s energy use.

Paralleling this, installing integrated HVAC systems to increase overall energy efficiency

could reduce use significantly. This would involve room-by-room use reductions, as mentioned

earlier, and could also involve more completely separating office HVAC from lab HVAC where

applicable. This might also involve integrating building heating and cooling into the campus

steam and chilled water loops, if any remain off of this system. This would be the most

expensive option compared with all others, and could even be prohibitively expensive to install

until buildings are torn down. Therefore, this recommendation should also be analyzed on a

building-to-building basis.

Technical Recommendations:

Because these technical recommendations involve new technology and installations, they

generally have a higher starting cost. They also may be restricted from implementation until a

new structure is built. Building managers may also argue against their implementation if the new

installations are deemed too costly to add onto the building. In the future, we hope that

14

additional technical advances will help reduce the energy use, but at the same time at a reduced

initial cost. When the initial cost is lower, more people are willing to implement the

technological changes because the change not as costly and at the same they will reap the

benefits of a cheaper energy bill.

Our first technical recommendation is to install sensor lights for office areas. Enough

sunlight enters the office space of the Vet Med building during the daytime, for several hours,

but lights continue to be kept on. This is because lights are turned on early in the morning when

there is not enough sunlight for the staff to see, but when there is enough light no one has the

mindset to turn off any unnecessary lights. This recommendation would allow for enough

sunlight to be in the room, but also make sure the room is lit enough. If the room lighting drops

too dim, lights will proceed to turn on again. Although this is implemented in the lab areas of

the Vet Med building, we hypothesize some of the sensors do not work, because the lighting

energy usage stays at a relatively constant rate even during the daytime, when the sunlight should

be brightest. Therefore, in addition to the installation of sensors, maintenance should frequently

check the lights to make sure they turn off at the correct time and that the sensors actually are

reactive to the light when the offices are brighter during the day. Also, maintenance should

check the efficiency and effectiveness of the light sensors, for example, if the light sensors only

turn off the lights when the room is extremely bright then maintenance can modify the sensors to

turn off the lights at a lower brightness scale.

Our second recommendation is a result of analysis of the electrical connection diagram.

The Vet Met 3B building has a built-in hookup for solar panels, but they have not been installed.

Because the hookup already exists, we assume this has been delayed due to cost of installation,

but in general solar panels are cost effective, and the input panel already exists, which further

15

lowers those initial costs. For older buildings, installing solar panels may be more complicated,

and therefore should be analyzed by each building to see what would be the best decision. In

general, if the solar panels are able to supply a good amount of energy to the buildings then it

should be cost effective to install them, and the payback will help give incentive to install the

panels. The initial costs would be higher for the other buildings compared to the Vet Med

building as the Vet Med already has built-in hookups but no panels, while other buildings do not

have built-in hookups. Solar panels have great potential in making a difference in the energy use

of Davis buildings because of the geographic location of Davis. The large amount of sunny days

will most likely make the solar panels cost effective at least in the buildings in Davis.

The first of these opportunities is installing roof temperature management panels to keep

buildings from heating during the summer. This will help the insulation of the building, keeping

the temperature more moderate and less susceptible to changes to outside weather. The structure

of the building should also be insulated as well to keep temperatures in the building constant.

Another recommendation for lab buildings would be to maximize useful lab space and minimize

wasted or unnecessary space. This recommendation will most likely need an architect and

engineer to help with building recommendations to calculate what the most effective and

efficient spatial use would be. Also, any space that is unnecessary will not be added to the

building so that the HVAC system, lighting, along with other machines that rely on the square

footage of the building. A third opportunity for lab buildings would be to keep office space at a

minimum and well separated from lab space. This is because the air from the labs needs to be

ventilated frequently and the office space does not need that much ventilation in comparison. By

keeping office space at a minimum, there is less space that needs to be ventilated which would

lead to less energy needed for the building. This recommendation is mainly for when new lab

16

buildings are built, as it is most likely too costly to change the set up of a building once it has

been built and in use. These recommendations are again limited to buildings that are not

currently as efficient as the Vet Med building.

*Need to note the idea of reducing kWh use factoring in to potential

savings???

Conclusions:

From our findings, we are able to observe various types of energy usages around the lab

buildings on the UC Davis campus. In the breakdown of energy uses, the HVAC is the most

energy intensive in all buildings. There are many ways to reduce energy use in the lab buildings,

but most are very minimal compared to any reduction in the HVAC energy use. One main goal

of our group is to change the mindset of workers in office and lab buildings. If they begin to

reduce their personal energy use, it will help policies change to reduce the energy use in

buildings in general.

17

Appendices:

MATLAB Code:

Data Importing and sorting:Note: % means comment lines%Import data into MATLAB, sorting rows as follows%alldata = [MSBAvg, MT1E, ME1ELTTEL, MT1EE1, MT2E, MT3E, MT3E1SPEC, % MT3E2SPEC, MT3E3SPEC, MT3E4SPEC, MT4, MT5E, MT6E, RMSBAvg, % MR1, MR2, MR3, MR4, MR5, MR6, MR7, MR8, MR9Spec]

daycount = 0;

sametime = zeros(22, 23); %Preallocate for speedweekend = zeros(8, 23);totaldayavg = zeros(96, 23);totalendavg = zeros(96, 23); %End preallocate for speed

for i = 1:96 %For 96 data points per day for j = i:96:length(MR1) %Taking all days at same time if (mod(floor(j/96), 7) ~= 2 && mod(floor(j/96), 7) ~= 3) %If day Not Sat Sun %Days start on Thursday, so Sat/Sun are mod(7) == 2, 3 daycount = daycount + 1; sametime(daycount, :) = alldata(j, :); %If true, take that whole row end end daycount = 0; totaldayavg(i, :) = mean(sametime); %Take average of weekday all points for time i and storeend

endcount = 0;

for i = 1:96 %Repeat for weekends for j = i:96:length(MR1) %Taking all weekend days at same time if (mod(floor(j/96), 7) == 2 || mod(floor(j/96), 7) == 3) %If day Sat/Sun %Sat/Sun are mod(7) == 2, 3 endcount = endcount + 1; weekend(endcount, :) = alldata(j, :); %Take whole row end

18

end endcount = 0; totalendavg(i, :) = mean(weekend); %Take average of all weekday points for time iend

Graphing Data:%alldata = [MSBAvg, MT1E, MT1ELTTEL, MT1EE1, MT2E, MT3E, MT3E1SPEC, % MT3E2SPEC, MT3E3SPEC, MT3E4SPEC, MT4, MT5E, MT6E, RMSBAvg, % MR1, MR2, MR3, MR4, MR5, MR6, MR7, MR8, MR9Spec]close allload('DataWAverages')

figure(1)hold on

%Plot total, HVAC, office receptacles, lab receptacles, total receptacles,%lighting, and HVAC broken outplot(Total, 'Color', 'Red')plot(MT5E+MT6E+MT2E, 'Color', 'Cyan')plot(RMSBAvg, 'Color', 'Green')plot(LabRecept, 'Color', 'm')plot(lighting, 'Color', 'Yellow')plot(OfficeRecept, 'Color', 'Blue')%plot(MT5E)%plot(MT6E)%plot(MT2E)

%Label axesxlabel('Time')ylabel('Instantaneous kW')

%Label curvesh = legend('Total power', 'Total HVAC', 'Total Plug Load', 'Lab Plug Load', ... 'Lighting', 'Office Plug Load');

%Set axes for time labelsset(gca, 'XTick', 1:500:2880)set(gca, 'XTickLabel', Time(1:500:end))axis([1 length(Total) 0 400])title('All data points without HVAC factor')hold off

19

%Plot weekday and weekend averages of abovefigure(2)hold onplot(dayMSBAvg+dayRMSBAvg, 'Color', 'Blue', 'LineWidth', 3) %Totalsplot(endMSBAvg+endRMSBAvg, 'Color', 'Red', 'LineWidth', 3) %Totals %To make legend work

plot(dayMT5E+dayMT6E+dayMT2E, '--', 'Color', 'Blue', 'LineWidth', 3) %HVACplot(endMT5E+endMT6E+endMT2E, '--', 'Color', 'Red', 'LineWidth', 3) %HVAC

plot(dayRMSBAvg, 'Color', 'Cyan', 'LineWidth', 3) %Total Recepplot(endRMSBAvg, 'Color', 'Magenta', 'LineWidth', 3) %Total Recep

plot(dayRMSBAvg - (dayMR1+dayMR2+dayMR3+dayMR4+dayMR5+dayMR6+dayMR7+dayMR8+dayMR9Spec), '--', 'Color', [0 1 0], 'LineWidth', 3) %LabRecepplot(endRMSBAvg - (endMR1+endMR2+endMR3+endMR4+endMR5+endMR6+endMR7+endMR8+endMR9Spec), '--', 'Color', [0.5 0 0.5], 'LineWidth', 3) %LabRecep

plot(dayMSBAvg - dayMT1E - dayMT2E - dayMT3E - dayMT5E - dayMT6E, 'Color', [1 0.5 0], 'LineWidth', 3) %lightingplot(endMSBAvg - endMT1E - endMT2E - endMT3E - endMT5E - endMT6E, 'Color', 'Yellow', 'LineWidth', 3) %lighting

plot(dayMT3E1SPEC+dayMT3E2SPEC, 'Color', [0.6 0 1], 'LineWidth', 2)

%Set axis, labels, etc.i = legend('Weekday Totals', 'Weekend Totals', 'Weekday HVAC', ... 'Weekend HVAC', 'Weekday total plug load', ... 'Weekend total plug load', 'Weekday Lab Receptacles', ... 'Weekend Lab Receptacles', 'Weekday lighting', 'Weekend lighting', ... 'Freezer Farm', 'Location', 'BestOutside');set(i, 'Color', 'None')xlabel('Time')ylabel('Instantaneous kW')set(gca, 'XLim', [1 96], 'YLim', [0 375])set(gca, 'XTick', 1:10:96)set(gca, 'XTickLabel', TimeNoDate(1:10:end))title('Average kW over all days at each time, weekday vs. weekend')

hold offhold off

20

grid on

%Caclulate percentagesfigure(3)percentHVAC = percentall(5)+percentall(12)+percentall(13);percentlighting = percentall(1) - (percentall(2) + percentall(6) + percentHVAC);percentoffice = sum(percentall(15:22));percentlab = percentall(14)-(percentoffice+percentall(23));FreezerFarm = sum(percentall(7:8));Vivarium = percentall(23)+percentall(9);labels = {'HVAC: '; 'Lighting: '; 'Lab Receptacles: '; 'Office Receptacles: '; 'Freezer Farm: '; 'Vivarium: ';};h = pie([percentHVAC percentlighting percentlab percentoffice FreezerFarm Vivarium], [0, 0, 0, 0, 0, 0]);

%% Following is to align pie chart labelshText = findobj(h,'Type','text'); % text handlespercentValues = get(hText,'String'); % percent values

combinedstrings = strcat(labels,percentValues); % text and percent valuesoldExtents_cell = get(hText,'Extent'); % cell arrayoldExtents = cell2mat(oldExtents_cell); % numeric arrayset(hText,{'String'},combinedstrings);newExtents_cell = get(hText,'Extent'); % cell arraynewExtents = cell2mat(newExtents_cell); % numeric arraywidth_change = newExtents(:,3)-oldExtents(:,3);signValues = sign(oldExtents(:,1));offset = signValues.*(width_change/2);textPositions_cell = get(hText,{'Position'}); % cell arraytextPositions = cell2mat(textPositions_cell); % numeric arraytextPositions(:,1) = textPositions(:,1) + offset; % add offset

set(hText,{'Position'},num2cell(textPositions,[3,2])) % set new position

%% Comparing with other buildings%Ghausi 67 Cooling 78 Heating kBtu/(sF*year)%Chem Annex/Chemistry 87 Cooling 98 Heating kBtu/(sF*year)%EPS 77 Cooling 88 Heating kBtu/(sF*year)%Vet Med 84 Cooling 68 Heating

21

Meter List:

22