TVE-E 19002
Examensarbete 15 hp Juni 2019
High resolution power measurement
Henrik Erlandsson
Johannes Eriksson
Jerker Ortman
Viktor Sköldheden
Teknisk- naturvetenskaplig fakultet UTH-enheten
Besöksadress:
Ångströmlaboratoriet Lägerhyddsvägen 1 Hus 4, Plan 0
Postadress:
Box 536 751 21 Uppsala
Telefon:
018 – 471 30 03
Telefax:
018 – 471 30 00
Hemsida:
http://www.teknat.uu.se/student
Abstract
High resolution power measurement
Henrik Erlandsson, Johannes Eriksson, Jerker Ortman, Viktor Sköldheden
In Uppsala there is a problem with overloading the electrical grid.
The cause of this is that too many consumers are using power at the same time. For example during lunch, when everyone is heating or cooking their food and big machines are still running. Because of this problem Uppsala has been denied installing charging stations for electrical buses. It also prevents the growth of the city. To find a solution the problem needs to be understood.
By measuring the consumption with higher resolution than today.
Patterns and deviations can be localized and solutions formed. In this project the focus has been measuring with higher resolution in different student accommodations for Studentstaden and sport facilities for Sportfastigheter.
Results shows that students living in dorms have a more well distributed power consumption compared to students living in
apartments. Although students in dorms are more likely to waste power by leaving electricity devices on, for example the lights in the
shared kitchen and the hallway. In order to solve this movement sensors can be installed and possibly giving the tenants more
information about the problem so they have the opportunity to change their everyday living behaviour.
As for Sportfastigheter who is a part of Live-in Smartgrid, which is a local power market place. Sportfastigheter needs to be able to lower their power consumption with at least 100kW on demand for at least one hour. The most promising solution for this was found in their cooling units for C-rink and Relita. By capping the cooling units to 70% of their maximum power the 100kW can be distributed. A more long term solution could be to connect the three cooling units in Relita to one. This could permanently lower their maximum power.
Uppsala University
Independent Project in Electrical Engineering
High resolution power measurement
Authors:
Johannes Eriksson Henrik Erlandsson
Jerker Ortman Viktor Sk¨ oldheden
June 20, 2019
Abstract
In Uppsala there is a problem with overloading the electrical grid. The cause of this is that too many consumers are using power at the same time. For example during lunch, when everyone is heating or cooking their food and big machines are still running. Because of this problem Uppsala has been denied installing charging stations for electrical buses. It also prevents the growth of the city. To find a solution the problem needs to be understood.
By measuring the consumption with higher resolution than today. Patterns and deviations can be localized and solutions formed. In this project the focus has been measuring with higher resolution in different student accom- modations for Studentstaden and sport facilities for Sportfastigheter.
Results shows that students living in dorms have a more well distributed power consumption compared to students living in apartments. Although students in dorms are more likely to waste power by leaving electricity de- vices on, for example the lights in the shared kitchen and the hallway. In order to solve this movement sensors can be installed and possibly giving the tenants more information about the problem so they have the opportunity to change their everyday living behaviour.
As for Sportfastigheter who is a part of Live-in Smartgrid, which is a lo- cal power market place. Sportfastigheter needs to be able to lower their power consumption with at least 100kW on demand for at least one hour.
The most promising solution for this was found in their cooling units for C-rink and Relita. By capping the cooling units to 70% of their maximum power the 100kW can be distributed. A more long term solution could be to connect the three cooling units in Relita to one. This could permanently lower their maximum power.
Contents
1 Introduction 4
1.1 Project Description . . . 4
2 Theory 5 2.1 Energy vs power . . . 5
2.2 Live-in Smartgrid . . . 5
2.3 The CoordiNet Project . . . 6
2.4 Hardware . . . 6
2.4.1 Electricity meter . . . 6
2.4.2 Eliq Starter Kit . . . 7
2.4.3 Three-phase measurement tool . . . 8
2.5 Databases . . . 8
2.5.1 ClimaCheck . . . 8
2.5.2 Eliq & Lifesmart . . . 9
2.5.3 Energiportalen . . . 9
2.6 Facilites . . . 10
2.6.1 Sportfastigheter . . . 10
2.6.2 Studentstaden . . . 10
3 Implementation 11 4 Results 12 4.1 Sporfastigheter . . . 12
4.2 Studentstaden . . . 16
4.2.1 Dorm rooms . . . 16
4.2.2 Apartments . . . 19
4.2.3 Laundry room . . . 21
5 Discussion 23 5.1 Studenstaden . . . 23
5.2 Sportfastigheter . . . 25
5.2.1 Insulation of the cooling units . . . 27
6 Conclusion 28 6.1 Studentstaden . . . 28
6.2 Sportfastigheter . . . 29
7 Acknowledgement 30
8 References 30
A Appendices 31
A.1 RTAC-200 . . . 31 A.2 RTAC CH530 . . . 37
1 Introduction
In recent time Sweden has been focusing on lowering the energy consump- tion, but barely any focus regarding power. This has led to lack of power supply from the grid to big cities, which is a problem for the expansion. All new companies and apartment builders are required to wait in line to be permitted their electricity needs, The municipality of Uppsala does not want to reject any builders but has in recent time been forced to reject both a bat- tery factory as well as a server building [1]. Since the focus has mainly been targeting energy, the measurements of electrical consumption are per hour.
The current measurement method gives a very vague picture of the electrical consumption pattern. This low resolution only provides data of the aver- age energy consumption by the hour but is missing the power consumption during this hour. Therefore it is impossible to localize the power pattern.
Without the power pattern it is much more difficult to find a solution to the problem.
1.1 Project Description
The purpose of this project is measuring the electrical consumption in student residences and sport facilities with a much higher resolution, analyze the data and finally try to come up with solutions to help solve the power issue.
As a part of Uppsala and as a big power consumer, Sportfastigheter is part of two projects called Live-in Smartgrid [2] and CoordiNet [3]. As a member of these project they have to be able to promise not to use 100kW of their power subscription on demand for shorter periods when needed, avoiding an overload on the grid.
• Is there any concrete ways to lower the power peeks in student facilities?
• Can Sportfastigheter fulfill the requirements from Live-in Smartgrid and lower their power usage by 100kW on demand?
• Is there any notable differences between high and low resolute mea- surements?
• Is high resolute measurements required to solve the problems above?
2 Theory
2.1 Energy vs power
Power is instantaneous which means if you switch on a lamp that needs 1000W that power needs to be delivered at that moment from the grid.
Energy measured in kilowatt-hours is the power over time. If the same lamp is on for one hour, the lamp has consumed 1 kilowatt-hour.
Figure 2.1: Power vs Energy
2.2 Live-in Smartgrid
As more micro power producers and renewable energy systems are entering the market Live-in Smartgrid have developed a test bench for innovative solutions for making the power grid more effective. The power grid as of now has bottlenecks were the power lines can not deliver the required power to sustain the expansion of cities. Big scale changes in the power grid is costly and takes a long time, therefor Live-in try to affect the power consumption instead. Internet of things or IoT for short, works as a sim card that can only be used to access the internet, meaning the cost of these cards are small and eliminates the need for installation of internet cables. With increasing numbers of units in facilities being connected to the internet, IoT has a
potential to work as the backbone for information sharing and regulating the power grid as part of Live-in Smartgrid test bench. [2]
2.3 The CoordiNet Project
CoordiNet is an EU project. ”The purpose of CoordiNet is to establish different collaboration schemes between transmission system operators (TSOs), distribution system operators (DSOs) and consumers to contribute to the development of a smart, secure and more resilient energy system.”[3] Uppsala is one of the ten test locations for this project. The project aims to establish cooperation between DSOs and TSOs to distribute the power usage more efficiently.
2.4 Hardware
2.4.1 Electricity meter
An electricity meter is an instrument installed in facilities to measure the consumed power. Old electricity meters had a display that was manually read and written down to keep track of the energy consumption. Nowadays the electricity meter has both a display as well as a LED that is blinking every kWh. The current and the voltage is measured and when the energy reaches a total of 1kWh the LED blinks. The blinks were measured with a light sensor that was taped onto the electricity meter. The power measured in the meter is divided by the number of measurements made in an hour and when the sum of each measured value reaches the value specified on the meter, the LED blinks.
Figure 2.2: Electricity meter
2.4.2 Eliq Starter Kit
A light sensor from Eliq was used to measure the impulses from the electrical meter. The sensor sends the impulse count wireless to the Smart Station.
The Smart Station calculates both the kWh and the power. The power is calculated by dividing kWh by the time between two impulses. The data collected by the Smart Station is sent to Eliq’s database via a GSM router connected to the station. [4] The Eliq Starter Kit was provided, so no other alternatives were researched.
Figure 2.3: Eliq Starter Kit
2.4.3 Three-phase measurement tool
The Three-phase measuring tool consist of three clamp meters, one for each phase. The measured voltage and current is converted and connected to an electricity meter. A light sensor is counting the impulses from the electricity meter similar to the light sensor from the Eliq Starter Kit to get both kWh and power.
2.5 Databases
2.5.1 ClimaCheck
ClimaCheck is a software that allows the user to monitor cooling, air condi- tioning and heat pumps energy consumption, efficiency and if the components are operating between set parameters. Sportfastighter had this software in- stalled in one of their hockey rinks and was used to get a approximation on which components were consuming the most power in the whole facility and at what hours power peaks were occurring.
2.5.2 Eliq & Lifesmart
The software used to connect and receive information from the sensors were from Eliq and Lifesmart. Lifesmart is the developer of the sensor and also the first place where the data can be accessed live. Eliq uses this data and forwards it to their own platform where the data is saved and stored into graphs for the user to see.
The Application Programming Interface API, can be accessed by obtaining a access-token from the website of Eliq. With the access-token data can be found with preferred date and time interval. The time interval can be chosen from 6 minutes, hour, day and data now. As seen in Figure 2.4 the data is displayed as a text and can from there manually be exported into another file where the data can be shown in a graph. For this project an already existing code were used to automatically download the data to Energiportalen for future use.
Figure 2.4: Example of live data from Eliq’s API
2.5.3 Energiportalen
Stuns has their own cloud where the data is saved for future use. This cloud stores all data that Stuns collects from different platforms as well as differ- ent Eliq and Lifesmart accounts. This data was saved by using the already known code from Stuns that over and over again scan for new data directly from the Eliq API.
2.6 Facilites
2.6.1 Sportfastigheter
Sportfastigheter consists of three ice hockey rinks and one bandy arena.
There is a total of three cooling systems. The A- and B-rink share one of these cooling systems consisting of one main cooling unit and two backup units. One of the systems consisting of a single cooling unit is used to cool down the C-rink. The last system is divided into three cooling units. This is because in the Relita-arena each unit cover one third of the ice field. For more details about the cooling system for the Relita-arena see A.1.
2.6.2 Studentstaden
Adress No of Apartments Avarage m2 ApartmentP erson Built Renovated
Dj¨aknegatan 5 4 65 2 1966-1967 1984-1985
Ekebyv¨agen 11-13 94 25 1 2002 -
Rackarbergsv¨agen 12-14 60 13 1 1964-1967 -
Rackarbergsv¨agen 46,48,52,54 146 14 1 1964-1967 -
Rackarbergsv¨agen 56 3 55 2 1964-1967 -
Flogstav¨agen 59 F 4 58 2 1971 -
Studentstaden 7 2 49 2 1952-1953 2000
Studentstaden 13 18 19 1 1952-1953 2000
Table 1: Information considering the facilities owned by Studentstaden [5]
One laundry room was also included in Rackarbergsv¨agen 12-14 were there was three washing machines, two tumbling dryers and one drying
cabinet.
All facilities in Table 1 consist of only student accommodations. Studentstaden 7, Flogstav¨agen 59F, Rackarbergsv¨agen 56 and Dj¨aknegatan 5 consist of apartments, all other addresses consist of dormitories.
3 Implementation
A number of electricity sensor with belonging Smart Stations from Eliq were installed at selected locations. The sensor was placed over the blinking LED on the electricity meters. Measurements were made using the sensor pow- ered from batteries connected wireless to the Smart Station. The station was connected to the internet via a GSM-router. Each blink on the electrical meter were registered on the sensor and sent to the station which forwarded it online to LifeSmart and Eliq online.
Each station could be connected to several sensors which resulted in low- ered costs as several sensors were to be installed in the same building. Each station was provided with a timer. The timer rebooted the station once a night to prevent a software error from Eliq. A total of twenty sensors were installed in nine different facilities.
A three phase measurement tool was installed in the cooling units belonging to the C-rink by surrounding all three phases wires inside of the machine with the clamp meters. The installation required an electrician.
4 Results
4.1 Sporfastigheter
As the ice rinks were used all day both during weekdays and weekends the results didn’t differ depending on the day. Therefore the plots below only shows data from a typical day.
Figure 4.1: Shows the power usage for the lights in the A-rink during 24 hours
Figure 4.1 The total power usage from the lamps in A-rink. The graph is taken directly from the preinstalled climacheck (see 2.5.1). As seen in the graph the lamps has different levels of light intensity. 100%,75%,50%,25%
and off.
Figure 4.2: Power consumptions of the C-rink in Gr¨anby during 24 hours
The data in Figure 4.2 is taken from Eliq’s (2.5.2) API. High resolution is average power every 6 minutes and low resolution is the average power every hour, which is the same as energy consumed in kilowatt-hours every hour.
Type High resolution [kW] Low resolution[kW] ∆P [kW]
highest peak 292,8 239,6 53,2
Power at greatest difference 292,8 239,6 53,2
lowest peak 155,9 151,1 4,8
Table 2: Data points from Figure 4.2
Figure 4.3: Shows one day’s power consumption of the cooling unit
The data in Figure 4.3 is taken from three-measuring tool (2.4.3). High reso- lution is average power every minute and low resolution is the average power every hour, which is the same as energy consumed in kilowatt-hours every hour.
As seen in Figure 4.3 the high resolution measurements shows that the cool- ing unit runs on ∼ 20% for about 30 minutes then runs on a greater load (from 30% to 100%) throughout the day. This information can’t be seen with low resolution measurements. For example during the hours 11, 12 and 13, the power consumption seems to be even with information from low resolution although it is clearly not.
Type High resolution [kW] Low resolution[kW] ∆P [kW]
highest peak 125,6 63,7 61,9
Power at greatest difference 125,6 63,7 61,9
lowest peak 49,2 37,6 11,6
Table 3: Data points from Figure 4.3
Figure 4.4: Efficiency of the cooling unit during one day
Figure 4.4 shows at what efficiency the cooling unit was running at during the same day as seen in Figure 4.3. During 48% of the time the cooling unit is running at its lowest efficiency.
Figure 4.5: Shows the power consumption when the cooler unit is
being rebooted
Figure 4.6: Shows the power consumption of the cooler during
normal conditions
unit for 10 minutes. The shutdown resulted in a different behavior than normal(Figure 4.6). The cooling unit turned on and off for several hours before going back to normal cycle. The turn on and off pattern resulted in greater power peaks than normal, although the energy consumption were almost the same.
4.2 Studentstaden
The data in the graphs below were chosen for one day. The power consump- tion slightly varies depending on if there is a weekday or weekend. Although the results were almost the same as students had no weekly routine in the same way as workers has. Therefore the plots were solely focused on a typical day.
4.2.1 Dorm rooms
Figure 4.7: Shows one day´s power consumption for one dorm house on Rackarberget.
The data in Figure 4.7 is taken from Eliq’s (2.5.2) API. High resolution is average power every 6 minutes and low resolution is the average power ev- ery hour, which is the same as energy consumed in kilowatt-hours every hour.
Figure 4.7 Shows the total power consumption of several student halls. The graph shows both the low resolution version (Per hour) and the high resolu- tion version (per 6 minutes).
Type High resolution [kW] Low resolution[kW] ∆P [kW]
highest peak 34,5 30,7 3,8
Power at greatest difference 33,6 25,5 8,1
lowest peak 11,3 10,7 0,6
Table 4: Data points from Figure 4.7
Figure 4.8: Shows the power consumption for one dorm house on Rackarberget over 24 hours
The data in Figure 4.8 is taken from Eliq’s (2.5.2) API. High resolution is average power every 6 minutes and low resolution is the average power every
Type High resolution [kW] Low resolution[kW] ∆P [kW]
highest peak 11,2 7,6 3,6
Power at greatest difference 11,0 6,5 4,5
lowest peak 3,0 2,7 0,3
Table 5: Data points from Figure 4.8
Figure 4.9: Shows the power consumption for one dorm house in Ekeby over 24 hours
The data in Figure 4.9 is taken from Eliq’s (2.5.2) API. High resolution is average power every 6 minutes and low resolution is the average power ev- ery hour, which is the same as energy consumed in kilowatt-hours every hour.
Seen in Figure 4.7,4.8 and 4.9 the pattern is the same as all of the plots that display student halls with dorms. Although the pattern looks similar a few differences can be distinguished. Firstly the newer house shown in 4.9
Type High resolution [kW] Low resolution[kW] ∆P [kW]
highest peak 7,7 4,7 3,0
Power at greatest difference 7,7 4,7 3,0
lowest peak 1,2 1,1 0,1
Table 6: Data points from Figure 4.9
4.2.2 Apartments
Figure 4.10: Shows the power consumption for one apartment in Flogsta over 24 hours
The data in Figure 4.10 is taken from Eliq’s (2.5.2) API. High resolution is average power every 6 minutes and low resolution is the average power every hour, which is the same as energy consumed in kilowatt-hours every hour.
Type High resolution [kW] Low resolution[kW] ∆P [kW]
highest peak 2,5 1,7 0,8
Power at greatest difference 1,4 0,5 0,9
lowest peak 0,2 0,1 0,1
Table 7: Data points from Figure 4.10
In Table 7, ”highest peak” is at the point where high resolution reaches its maximum during the day, ”greatest difference” is at the point where the difference between high resolution and low resolution is the greatest and
”lowest peak” is at the point where low resolution is at its minimum, and the high resolution value is the peak during that hour. ∆P is the difference between high resolution and low resolution.
Figure 4.11: Shows the power consumption for one apartment on Rackarberget over 24 hours
The data in Figure 4.11 is taken from Eliq’s (2.5.2) API. High resolution is average power every 6 minutes and low resolution is the average power every hour, which is the same as energy consumed in kilowatt-hours every hour.
In Figure 4.10 and Figure 4.11 two different student apartments are shown.
In both these graphs it is very clear when the residents are home.
Type High resolution [kW] Low resolution[kW] ∆P [kW]
highest peak 3,1 1,2 1,9
Power at greatest difference 3,1 1,2 1,9
lowest peak 0,1 0 0,1
Table 8: Data points from Figure 4.11
4.2.3 Laundry room
Figure 4.12: Shows the power consumption for the laundry room over 24 hours
The data in Figure 4.12 is taken from Eliq’s (2.5.2) API. High resolution is average power every 6 minutes and low resolution is the average power every
The plot in Figure 4.12 displays the power usage in one of the common laundry room located in Rackarberget. The high resolution shows that the power peaks is high compared to what is shown in the low resolution graph.
As seen in the graph the laundry machines can be accessed from 8 am to 22 pm. The power usage outside of these time frames were from the tumblers and the drying cabinet.
Type High resolution [kW] Low resolution[kW] ∆P [kW]
highest peak 17,9 8,2 9,7
Power at greatest difference 16,3 6,1 10,2
lowest peak 0 0 0
Table 9: Data points from Figure 4.12
5 Discussion
All the electric meters were located at a position that made it hard for us to supply our sensors with an internet connection. This problem made us decide that we would use routers provided with SIM-cards. Unfortunately we couldn’t get hold of enough routers to start our measurements on all the facilities. Another problem with the installation were that the Smart Station from Eliq seemed to have a software bug. The station did shut it self down with an interval of a couple days. At first we were told to just go and restart the Smart Station as well as update the system. However it happened again so instead of manually restart the stations every other day timers were installed. The timers would restart the station every night for fifteen minutes. This would give us a gap in the information gathered by the sensor but was a necessity.
Due to delivery delays and limited time, measurements were only made for about two weeks time. Although these days gave similar result it would be interesting to have measurements for at least a year. Obtaining a year of data would make it possible to compare the power usages for different seasons.
As there are two companies affiliated with this project our main focus was towards Sportfastigheter. The reason for this being improvements regarding power consumption and making their facilities more effective was to be our priority. It was easier to measure subsystems such as cooling, ventilation etc on their facilities rather than Studentstaden were there was only possible to measure the apartments and one laundry room, resulting in having more opportunities for improvement on Sportfastigheter than Studentstaden.
5.1 Studenstaden
The graphs of the dormitories have their distinct look due to a shared kitchen in each hall, cooking is protracted resulting in even power usage instead of huge power peaks during dinner time. The students living in dorms had the electric bill included in their rent, a side effect of this as seen in 4.7 is that the power never drops low. The main reason for this is lights never being turned
in their rooms. Not to forget the fridges and refrigerator’s in the kitchen is always turned on and is the main reason to the power never dropping to zero.
As in the dormitories the apartments also have an distinct look but with a lower power between the peaks. Because the numbers of residents is different in each accommodation the power level between the peaks is difficult to interpret. However the students living in apartments pay their own electricity bill, so they might be more aware of their power consumption. Also in dorms the light in the kitchen and the hallway is not effecting the student in the same way as in the apartments. They also have more apparent peaks. Usually the peaks occurs during some kind of meal, such as breakfast, lunch, and dinner.
The sources to the peaks is machines that drain a lot of power such as stove, oven, kettle, vacuum cleaner, microwave etc. The power consumption used when the apartment is empty is most likely the refrigerator and the fridge.
We believe that the power consumption could be reduced by including the residents in the solution by showing all the people living in Studentstadens estates what they and their neighbours are consuming. Hopefully the social pressure would make the residents think of their power usage. To make it even more thrilling we suggest that a weekly, monthly, or yearly prize to the most efficient household in each of Studentstadens residential areas are awarded. All this could be presented for example on an application showing the power usage. However there would be no need for this measurements to be highly resolute as it is only the total consumption (in a period of time) that is relevant.
Our results indicate that the power usage is higher in the dorm rooms com- pared to the apartments during hours when people normally are at school.
As a solution we suggest Studentstaden install motion detectors in all their facilities and timers in the kitchen and showers to reduce unnecessary power usage when people are not at home. But if we take in consideration how many residents there is, the power usage per person is lower in the dorm house.
A solution to the overload on the grid would be disabling the common laun- dry room during the most power consuming hours which is lunch and dinner
5.2 Sportfastigheter
By swapping out the lights that are currently in the A, B- and C-rink for LED lights the power peaks would not decrease, but the over all power con- sumption would drop as LED lights are more power efficient. Doing this would mean that all the armatures would have to be replaced resulting in a discussion of cost efficiency. This has not been done in our case. However we still suggest swapping to LED lights as they are better in the long term.
With the current setup for cooling the C-rink and Relita it is rather inef- ficient. A short term solution could be by combining the cooling loops for C and Relita and thereby increasing the load for the units and improving the efficiency of the system as seen in appendix A1. As a safeguard one addi- tional unit could be installed to act as a backup if the other two fail to cool the ice.
A more long term solution would be to replace the current cooling units with newer one and repeat the short term solution as the current units are getting close to the end of their life span.
To be a part of Live-in Smartgrid, Sportfastigheter must at any time be able to lower their power usage by a minimum of 100KW. They must be able to do it for at least one hour, so that their power capacity can be redistributed to relive stress from the power grid. This could either be 100KW from the roof of their power subscription or from the current power usage. The cool- ing units ”on and off” behaviour is the biggest problem in the pursuit of the 100kW goal. This is seen in Figure 4.3. To pass over 100KW from the subscription roof is not a problem if the suggestions above are implemented.
However if the reduction needs to be from the current power usage it will cause problems. A solution for this could be to use a buffer for the coolant.
When power needs to be redistributed by Live-in Smartgrid, Sportfastigheter could use the buffer with the cold coolant and shut down the cooling units to eliminate these power peaks. This buffer needs to be large enough to ensure that the ice stays cool for at least one hour as Live-in Smartgrid have these demands. When the redistribution is no longer necessary, the cooling units can be turned on again and begin cooling the ice as well as the buffer. No
needed to determine if this is a valid option.
The subscription from Vattenfall allows 1650kW. This subscription includes Relita, C-rink and the track and field arena. Since no information about the total usage from the track and field arena is available, only savings from the cooling units are taken into consideration.
The total usage from the cooling units are estimated to be four times the usage from the cooler on the c-rink (the c-rink is estimated as one third of Relita). The highest power used by the coolers were measured to be around 165kW when rebooting and around 140kW in normal drive. The cooling units are programmed to always run in normal drive and never to reboot.
This means that if there is no blackout, the total highest usage from four cooler units is around 600kW.
The coolers are implemented with a control system CH-530 (appendix A.2).
In the control system there is settings to lower the load. This means that instead of going to 140kW the cooler can be told to stop at lower load for example 100kW ( 70%). An adjustment like this will save a total of 40*4=
160kW on demand. A decrease in load could also be a permanent solution to lower the subscription. To be able to see if this method is viable, the system has to be tested for a few weeks. The test would see if the cooler units are able to maintain the same ice quality at 70% of maximum capacity.
The main issue with this is that if there is a blackout or another reason for the cooler units to shut down, the coolers has to be turned on one at a time to avoid the subscription limit to be exceeded (compare Figure 4.3 and Figure 4.5).
Without researching the background of the subscription and possibly already implemented improvements, the subscription seems a bit too generous. The subscription consists of Relita and C-rink that is mainly being used on the winter and the track and field arena that is mainly being used on the summer.
These two only collide during the spring and the autumn and the coolers for Relita is partly off during both the spring and the autumn.
5.2.1 Insulation of the cooling units
As of now the cooling units that handle the C-rink and Relita are located in an uninsulated building. Not only does this effect the cooling system in a negative way by making them more inefficient. They also stop working when the outdoor temperature is below -18°. As the chiller only works between -18°in the low ambient option and 52°in the high ambient option, the staff has to manually install temporary heating fans to heat up the chiller during the colder days of the winter. Not only is this a waste of energy but also effects the grid since the winter is the time of the year when the grid is at a critical level. The solution for this would be to build the chiller into an insulated house with an opportunity to manually ventilate. This would keep the heat during the winter and let the heat out during the summer.
6 Conclusion
Even if a few smaller consumer’s would change their habits and lower their power peaks, it would take quite a lot of people to make a difference. If bigger consumers such as Sportfastigheter were able to lower their total con- sumption or even on demand lower their consumption by 100kW it would benefit the community. Not only would they help the grid but also lower their bill costs. This is not only possible but there is also solutions ready to be tested.
The project was limited by the short span of time the sensors were out in the field. Also a big limitation were the restriction of not being able to visit the facilities whenever needed but instead had to wait for the janitors to have free time in their busy schedule and the lack of knowledge from the facility owners and janitors. There were no one information about exactly what dorms and houses each of the electricity meters were measuring
6.1 Studentstaden
For Studentstaden and their tenants it is harder to drastically lower their consumption, especially the power peaks. But what they can do is to lower their idle power consumption. This by implementing movement sensors for the lights in all their buildings as well as timers on the stove and oven. In the apartments the only easy way to lower the power consumption is to replace the old machines with new more efficient ones.
Dorms had a more even power distribution than apartments but the apart- ments had lower power usage at their minimum. This is probably a side effect of students in dorms are less aware and more careless if the lights are on in the shared kitchen or if they install their own refrigerator because it’s comfortable to have one in their room. Also the newer buildings have a big- ger difference in lowest and highest power consumption over a day. This due to newer and smarter technology. Regarding high resolution measurements it is not necessary to monitor the consumption but gives a more precise and fair image. It would also be interesting to see how the power consumption differed in different seasons.
After comparisons between apartments and dorms were made, it is clear that in the eyes of power consumption, it is better to live in a dorm. This because when the power peaks gets distributed over a longer time period instead of pack together at the same time, it will not affect the grid as much. At the same time there is a bigger opportunity for improvements in dorms.
Future research for Studentstaden could be the affects of making people more aware of their usage and ad a social pressure to the mix or why people want to live in apartments rather than dorms.
6.2 Sportfastigheter
The most important finding in Sportfastigheter were the fourteen year old cooling units, since they were initially made for the bandy field when it was outdoor, but later a shell was built around it. The cooling units could be more efficient and less power consuming if they were connected together. When the study was done the cooling units was during 50% of the time running on low power (∼15-20%) but if two of the cooling units got connected together at the same time as a third one is backup the units would run on higher power which would result in higher efficiency (Appendix A.1). Also the lightnings were recalculated and if the old lightning is changed into newer ones the power would also decrease.
Sportfastigheter should look into the possibilities to use some kind of buffer to ensure a future cooperational work with Live-in Smartgrid. They can also look closer into what the track and field arena consume and when the total consumption is at its peak. With the purpose of lowering the electrical sub- scription from Vattenfall.
7 Acknowledgement
Joachim Lindborg, Rafael Waters, Sara W¨angelin, Mikael Larsson, Karolina Gahne, Herfinn Olsen, Jonny Johansson,
8 References
[1] Lindblom Mari, U N T , U ppsala har slagit i eltaket, 2018-12-10
[2] Publisher Lindborg, Joachim, https://www.live-in.se/om-live-in, 2019- 05-10
[3] Uppdated 2019 https://coordinet-project.eu, 2019-05-02
[4] Clark, Tom. 2017. T eknisk specif ikation Eliq Online. Eliq https://eliq.zendesk.com/hc/sv/articles/115002726505-Teknisk-
specifikation-Eliq-Online, (H¨amtad 2019-05-10)
[5] Updated 2019, Publisher Nilsson, Christian
https://www.bostadsuppgifter.se , 2019-05-01
A Appendices
A.1 RTAC-200
A.2 RTAC CH530
Gränssnitt för styrfunktioner
Börvärdesskärmbild
Börvärdesskärmbilden är indelad i två delbilder. I delbild 1 förtecknas alla ändringsbara börvärden tillsammans med sina aktuella inställningar. Operatören väljer vilken parameter som ska ändras genom att beröra beskrivningen eller värdet. Om du gör det växlar skärmen till delbild 2.
På bild 1 är börvärdet för språk alltid lika med det sista på listan. Detta underlättar språkändring genom att den angivelsen finns i en och samma position i alla CH.530- aggregat.
Bild 2 visar på sin övre halva aktuellt värde på det valda börvärdet.
Det visas i ett ändringsbart format som är anpassat efter dess typ.
Binära börvärden ses som en enkel tvålägesangivelse för vilken alternativknappar används. Analoga börvärden visas med
rotationsknappar. Bildens undre del används till hjälpbilder.
Modes Chiller Compressor
Auto Stop
Auto Local or Remote:
Front Panel Chilled Water Setpoint:
Front Panel Current Limit Setpoint:
Condenser Limit Setpoint:
Low Ambient Lockout Setpt:
Low Ambient Lockout:
Local 7 C 100%
XX % HPC 1,7 C Enabled
