PRE2023 1 Group1

Group members

Name	Student number	Major
Sven Bendermacher	1726803	BAP
Marijn Bikker	1378392	BAP
Jules van Gisteren	1635530	BAP
Lin Wolter	1726927	BAP

Problem statement

With the emerging of more renewable energy and of more energy intensive technologies, like airconditionings needed due to rising temperatures and more extreme weather, the electricity landscape has changed. The increase in load on the electricity net has to be counteracted by either improvements of the network or measures to decrease the peakload on the net. Next to this consumers often pay high prices for electricity due to producers having to power on fossil fuel plants to reach the electricity production needed for the peak times of the day, while sometimes renewable electricity is lost due to lack of use during high sun low use times. To fix this problem producers and governments have turned to introducing dynamic electricity contracts, where consumers pay different amounts of money depending on the time of consumption, thus making it more financially wise to use electricity on the low peak times. This usage of the electricity on different times can even lead to profits on the end of the consumer, and helps the planet and producers to increase the share of renewable sources of electricity. The dynamic electricity contract needs, at the moment, still a large amount of effort and input from the consumer to achieve the promised savings on electricity bills. Which thus leads to less people switching over to a dynamic electricity contract hampering the decrease of load on the net and inclusion of renewable electricity sources. A system is thus needed to decrease the effort needed to be done by the consumer, such a system would in the best case automate all appliances based on the before set operation times and the known electricity prices. This automation would need a house full of smart appliances, which are connected to the internet and can thus be turned on remotely. Research done into these systems has shown that consumers need direct messages in order to be reminded of the time to turn on the appliances, which then leads to a efficient use of the dynamic electricity contract, but this research has also shown that consumers would rather have everything automated. This automation however is still very hard due to the lack of smart appliances in most households. Research also shows that consumers benefit from an overview of their electricity usage. The system to solve this problem would thus need at minimum an overview of electricity usage, a time window input, to determine the times the consumer would accept and be able to turn on the appliances, a message that is send at the prime-time the consumer has to turn on the appliance which could include the savings achieved by the to be undertaken action and in best case a direct link to the smart appliances to automate the activation of the appliances.

Furthermore, if the automated connection of smart appliances is not yet a possibility to be developed or the user still has appliances which are not smart the info about the appliances is still needed. In order for the system to work well the input parameters of the appliances is needed as well as the information about the length of washing programs or other time related variables. Without this data the system would not be optimized, since it would have many unkowns that would lead to problems for the consumer. With this data however the best time to turn on the appliances can be chosen thus leading to no lapses in judgement by the system. It is for example important that when a program takes 3 hours the to be done task by the appliance is completed on time before the maximum finish time set by the user. Another important factor to take into account for the system is that solar panel users need other advice than people without solar panels, since the solar panels can sometimes cost money, due to negative prices, or can save much money due to times of high solar intensity. The negative prices are caused by overcongestion on the electricity grid, which if not handled would lead to failure of the grid, this is especially caused by the large amount of renewable, unstable sources of electricity. Even the minister of energy of the Netherlands states that it is sometimes more beneficial for consumers with solar panels to turn them off in times of high negative prices, this is something that the system could also control or atleast mention to the consumer. In times of no negative prices the use of solar electricity is always free, thus the system must take this into account. The price of electricity might be lowest in the morning but when the solar intensity is higher in the afternoon it would be more cost benificial to turn on the appliances during times that the consumers' solar panels are peaking in electricity generation. One final part of the problem is the possibility for people to use their electric cars as electricity storage, for this to work however many things have to be kept in mind. The electric car has to still be charged for the travel to be done by the user, thus another input is needed where the minimum charge can be set by the user, this in turn then enables the system to better optimize electricity use since the car could be charged at times of low prices and then discharged at peak times thus leading to even more savings for the user. The health of the car battery however has also to be taken into account, since the savings on electricity bills would need to counteract the degredation of the batteries well enough to be worth it. This is dependent on the age of the car as well as the type of car. The system would thus need much more info and should in the best case be directly connected to the software of the electric car.

The system that is then developed has to be easily accesible for users, this could be either done with a website or an app, the app will probably lead to the best possible integration since the messages that have to be send would in case of the website require the phone number of the user. This app would then contain options to give all information and inputs stated above to be required, able to be put in automatically with a connection to a database, for example of appliances, or manually by the user. The app would then also be connected to the other systems the user has such as their electricity providers app, the app of smart appliances and app of the electric car. With all this the system, most likely in app form, would then become a hub for all the users electricity based needs. Where the system even optimizes dynamic electricity pricing to be as profitable as possible, in turn helping the electricity network by deloading the network at peak times enabling the better inclusion of renewable forms of electricity.

Hypotheses

To see whether our product has solved the problem stated hereinabove we constructed several hypotheses, each of which states a different prospect about the effect of our product. The hypotheses are as follows:

1. Our product will keep electricity costs of households with a dynamic contract below the electricity costs the same households would have with a variable contract, assuming the contract is started at the time of our project.
2. Our product will keep electricity costs of households with a dynamic contract below the electricity costs the same households would have with a fixed contract, assuming that the contracted is started at the time of our project.
3. Our product will reduce the network congestion caused by households.
4. Our product will reduce the green energy wasted due to differences in supply and demand on the energy network.
5. Our product will make sure households will make use of green energy rather than fossil fueled energy.

Approach

On this wiki we will document the different ways in which these hypotheses will be tested. The first two hypotheses will be tested by comparing the energy costs of a consumer using our product with a dynamic to the energy costs of the same user having either a variable or fixed contract. This will be calculated for different scenarios to get a concise outcome of these comparisons. Hypotheses number three through five will be tested by performing literature study and interviewing involved parties.

Deliverables

The deliverables of this project will exist of a mobile phone application which will tell users when to turn on their electric appliances in order to get the lowest possible energy cost. The app will achieve this by checking for the dynamic energy pricing, having knowledge of all the electirc appliances available and calculating for each appliance when it should be turned on, keeping in mind the duration that a device needs to stay on. To properly keep track of what requirements the app needs to fulfil, what requirements we would want it to fulfil and by what the developement of the product will be constraint by we constructed three lists in which these points will be adressed.

Requirements

• The product needs to be connected to the internet in order to know the energy prices of the coming twenty four hours;
• The product needs a list of electrical devices along with information about the manner in which they consume energy (i.e. a dishwasher uses electricity in a cycle of a few hours, while a freezer needs to be on the entire time);
• The product needs an algorithm that optimizes the time of use for each device in its device list;
• The product needs an environment which the user can interact with (e.g. an application on a mobile device, a website or a device of its own);
• The product needs to be able to send information to the user by either notifications or a screen, or the product should be able to send signals directly to the devices which should be turned on/off;
• The user needs to be able to input information about their preffered time of receiving notifications (should this be applicable);
• The product needs to show the money that was saved in comparison with a variable electric contract;
• The user should not feel disrupted by the app (e.g. the app should not send notifications that would disrupt the user);
• The environment in which the user interacts with the product should be easy to navigate, such that users with every level of experience with technology should be able to know what they are telling the appand what the app is telling them (e.g. old people, who generally have a lower level of experience with technology should not be confused by the environment).

Preferences

• The electirc devices for which the product will optimize their electricity use should all be smart devices, such that the user does not need to turn on the devices manually;

Constraints

• The great majority of houses are, as of now, not (fully) equipped with smart devices, which means that our product should send notifications to the user telling them the time at which they should have the electric devices turn on;
• The users will not always be close to their homes and their devices at the optimized time for their devices to be turned on. This leads to multiple constraints: the user needs to have their devices should have prepared all devices in advance before they go away from home; The devices should all either have an option to schedule the time they turn or should be smart devices.

Users

The possible users for the algorithm for low energy pricing are quite vast, ranging from private homeowners to businesses. Private homeowners can use this app to lower their expenses on energy, which is especially important due to the surge in energy prices due to everything happening in geopolitics. Homeowners could thus use this to turn on their appliances at the right times leading to huge chances on saving large sums of money. Next to private homeowners even factories or businesses could look into using the algorithm, their sometimes-intensive use of energy could then also be better placed at more beneficial times. Energy intensive procedures needed to for example fabricate a certain product could then be done at better times lowering the costs of production leading to higher profits, which is in capitalism of course one of the main drivers in business. Thus, the users for the algorithm spread almost everyone, since almost no one lives without using electricity in this current era.

Private homeowners

As described above one of the most important users would be private homeowners, since the developed app/algorithm would enable them to save large sums of money. This users most important wish would encompass mostly a good working app which is easy to navigate as well as a trustable algorithm. The algorithm should be trustable since if the algorithm is wrong very often there would be no need to use the app, some errors can be accepted though since in no case would using the app cost more money than when using electricity on chosen times, only a perfect human being would be able to spread the electricity usage better than the algorithm.

Companies

The companies would similarly to the homeowners also want a good working app, with again a huge focus on the accuracy of the algorithm. Companies would also benefit from other built in functions such as overviews of electricty use as well as a method to make sure the responsible people are informed, which should thus need everyone to be connected to a single account. Companies would thus make profit from the use of the app/algorithm but would not need major alterations to the normal version for private homeowners.

Other institutions

Since the use of electricity is such a general thing in life, any group of people, company or institution having control over their electricity use could use the app to lower their electricty prices. An important question that then arises however is, in the case that many people start using the app, can the algorithm adapt to correctly spread the use of electricity. Since the increase in usage would also lead to an increase in electricity use on times that would otherwise be seen as off peak.

Other uses

Next to the intended uses done by users as desribed above it is important to mention that with a app based on a great algorithm other interests could also arise. If the algorithm is made to be very accurate with forecasts further into the future, as little as one day, people could start to make money using the app, since some platforms may enable users to trade electricity. This trading of electricity would increase prices and is in the scope of the perceived app not to be wished, since the proposed function of the app is to reduce costs and increase sustainability not to create profits for certain people.

Electricity companies / Network operators

The electricity producers and network operators are also users that have to be taken into account when it comes to the app. While they may not be direct users of the app it is in their best interest to have as many consumers use the app since this would lead to an increase in dynamic electricity pricing users and make the existing users more efficient. This would then lead to a decrease in load on the net on peak times and a more even spread which is positive for the network operators. The electricity companies would also have to spend less on expensive sources of electricity such as gas to fullfill peak demands since cheaper forms of electricity like solar and wind would be better used during low peak times.

User interviews

Interest interview

At the start of our project we interviewed a user of a dynamic contract, which can be found in Appendix B.2. Before this interview we let the interviewee sign an informed consent about the interview, which can be found in Appedix B.1. In this interview we found that he would be interested in the product, that it could be a problem that devices should be prepared at all times to be turned on at the right time and that he uses a dynamic contract because both economic and environmental reasons. We realized that we should be more specific in our interviewing style however, after we gained some more information on our product - by literature study, brainstorming and working towards a usable product -, which is why we only conducted this interview once

Scenarios

In order to get a better view on the typical electricity usage of possible users of our product we sketched two different scenarios: A single family household consisting of two adults and one child, and a single person houshold.

Single family household consisting of two adults and one child

They turn on the dishwasher, dryer and washing machine 1 time per day. They have no electric car and no solar panels. Between 20:00 and 1:00 the TV is on. From 21:00 to 24:00 the child runs a Playstation 5. The freezer and fridge are on at all times. There is no miscellaneous electricity use in this perfect scenario.

A fixed contract would at the moment cost around 0.345 euro per kWh, a variable contract would cost around 0.37 euro per kWh. For the dynamic contract the prices of September 28 were used, these prices are in comparison to other days quite poor for dynamic contract consumers.

Let's say that the household has appliances that are quite average when it comes to electricity use. The Samsung dishwasher is always run on the eco program leading to a use of 1.053 kWh for 195 minutes, where we will assume that this is spread evenly over the time. For the dryer it is assumed that a ‘mix’ cycle is always used and that the data from 2016 is accurate for the dryer owned by this household. Thus, the dryer will use 0.66 kWh over an assumption time span of 2 hours. For the washing machine it is assumed that 0.8 kWh is used, spread also over 2 hours. For the TV it is assumed that it uses 0.2 kWh per hour. The PS5 with TV combo uses 0.3 kWh per hour. The fridge uses 0.015 kWh per hour and the freezer also uses 0.015 kWh per hour. For the dynamic contract it is assumed that the appliances are turned on at the best possible times and that the electricity use of appliances is spread evenly over the time.

Adding the entire electricity usage up found is a total use of 5.11 kWh. Thus, the cost of this electricity usage can be easily calculated for the fixed and variable contract. With a fixed contract this electricity use would cost the household 1.76 euro and with a variable contract it would cost 1.89 euro. For the dynamic case the calculation is a bit trickier since the price is different for every hour. In order to spend the least amount of money with the dynamic contract with the electricity prices of 28 September the dishwasher should be turned on at 02:00 thus it is assumed that the appliances in this case are ‘smart’ enough to be turned on automatically. The dishwasher uses 0.0054 kWh per minute. The costs of running the dishwasher from 02:00 to 05:15 is then 0.279 euro. The cost of the washing machine turns out to be 0.208 euro, for the dryer 0.172. The TV is run during fixed times and would cost for this specific day 0.332 euro. The PS5 and TV combo would cost 0.297 euro. For the fridge and freezer, the total cost is 0.228 euro. Thus, the total cost with the dynamic contract is 1.52 euro. This would mean that on a very average day the savings are 0.24 euro in comparison to a fixed contract and 0.37 euro in comparison to a variable contract. Which, assuming this is an average amount of savings, would save this household 87.6 euro on year basis comparing to a fixed contract and 135 euro comparing to a variable contract.

Single person household

This person lives alone in a small apartment, the dishwasher, washing machine and dryer are turned on once every 3 days. This person also owns an electric car which is used for commuting from and to work, needing to only be charged a small amount at home since most of the charging is done while the person is at work. In this case the electric car is a Tesla model 3 which needs to be charged 20 percent during the night. The person also powers on a TV from 20:00 to 24:00 and cooks using induction for 30 minutes at 18:00. The electricity pricing data from Oktober 1 is used where it is assumed that today is the day the person turns on their appliances.  

The person uses a washing cycle costing 1 kWh spanned over 2 hours; the dryer is put on the ‘mix’ cycle and uses 1 kWh since this dryer is quite old and thus less energy efficient. For the dishwasher the quick cycle is used, using 0.75 kWh in 30 minutes. Charging the Tesla 20% will cost 10 kWh, which takes approximately 1 hour. The induction cooktop will use 0.75 kWh. Finally, the TV will use 0.15 kWh per hour.  

For the fixed contract a price of 0.345 euro per kWh is taken and for a variable contract 0.37 euro per kWh. The total kWh consumption is 14.1 kWh, which costs with the fixed contract 4.87 euro and with the variable contract 5.22 euro.  

Choosing the best time for the dynamic contract we assume that the appliances are ‘smart’ and can thus be turned on at all times during the day. The charging of the car however has to take place during the night. The appliances are turned on at 13:00, leading to a cost of 0.51 euro for the dishwasher, dryer and washing machine. The cooking would cost 0.25 euro and the TV would cost 0.2 euro. Finally, the charging of the car will be done at 03:00 leading to a cost of 2.9 euro. This is in total a day cost of 3.86, which saves 1.01 euro in comparison to the fixed contract and 1.36 euro in comparison to the variable contract. This can however be highly influenced by geopolitics and the weather.  

State-of-the-art

To get a firm grasp of our subject we conducted literature research by reading numerous articles related to the subject. We documented our research below in the form of a short summary of each article respectively.

Research on consumer risks and benefits of dynamic electricity price contracts

This research states that there is serious risk involved in switching from fixed prices to dynamic prices. It concludes that there is only little room for flexible electric consumption, and that the average dynamic prices in France and Austria were higher in 2021 than the fixed price. Furthermore it is said that for households with an electric vehicle(EV) a dynamic electricity bill could be beneficial, since a EV is the biggest consumer within the flexible electricity consumption activities. However, in this paper it is stated that most of the electric consumption is used for space heating and water heating. In the Netherlands we use natural gas to warm up our homes, so the situation might be different and more profitable for Dutch households.

Asset Study on Dynamic retail electricity prices

This research says that consumers can significantly decrease their electricity bill by shifting to low price moments. It evaluates different kinds of dynamic pricing options, Real time pricing, time of use up till critical peak pricing. Real time pricing and time of use pricing are the riskiest but yield the highest reward, the critical peak pricing has the lowest risk but yield a lower reward. This research also states that a dynamic pricing leads to a more efficient electric grid, since lower peak demand reduces the losses in the electricity grid. This also results in a lower electricity bill.  Additionally, dynamic prices incentivise demand shifting to times of lower prices which usually indicate times of high intermittent renewable energy resources (RES) feed-in. The use of excess electricity can reduce local congestion and therefore facilitate the integration of RES in the energy system. Therefore it would also be in the interest of the government to promote switching to dynamic prices.

Furthermore it gives an overview of the potential customers for dynamic pricing. With a premium on the electricity prices reducing the maximum prices, up to 90 percent of the costumers could profit from dynamic pricing.

Dynamic electricity pricing — Which programs do consumers prefer?

TOU vs RTP, RTP is real time pricing where the user pays based on the real time market prices which change every hour. TOU is time of use pricing where the price is fixed long in advance on a timetable.  

Consumers are fine with using dynamic electricity pricing as long as their daily routine is not affected by it or lead to reductions in their comfort level. When asked about electricity contracts people seem to still prefer a static contract. People tested in a dynamic pricing situation proposed multiple insights. Things like lights, tv, stove were things where the price at that point was not really taken into consideration, since the people thought it would affect their lifestyle too much. Things like dishwashers, washing machines and tumble dryers were used much more at low price times. Due to work however the participants of the test could not always benefit from the low prices and seemed to be less willing to turn on the appliances very early or very late in the day. People also preferred RTP over TOU since RTP made the users feel that they could save more money.  

Via a questionnaire it was found out that most people prefer a constant rate even though the advantages of a dynamic contract were explained thoroughly. The cost-saving expectation was 50 – 150 euro but turned out to only be 20 to 60 euro. Therefore, it is necessary that the other advantage of dynamic pricing, the load shifting, to be explained thoroughly. And the participants expressed a wish for demand automation, thus turning on the appliances at low price times automatically in order to make the dynamic pricing seem worthwhile.

Influencing residential electricity consumption with tailored messages: long-term usage patterns and effects on user experience

Persuasive technologies are an important method to alter consumer behaviour next to financial benefits. Personalized persuasive technologies work better to stimulate people into doing what is wanted of them. Things like tailored information, personalized content, cooperation and competition are known to be good design principles. Messages to the user should be send at appropriate times and should be managed to not create irritation.  

In a trial done with real household many insights were found by the authors. The program where users could gain more insights next to only SMS messages stating the best time to turn on the appliances was used quite little by the users, the users that did at first often use it proceeded to use it less as the trial went on. The users were happy with being able to see achieved savings as well as being able to see a comparison between real and optimal consumption curves. Users say that an in-home display could have stimulated them even more. Also, next to time and possible savings the rate should also be included in the message sent to the users. It was seen that the washing machines and dryers were shifted most often to accompany electricity savings, while dishwashers the least. Users again brought up a preference for automation. Users finally also mentioned that while they were willing to alter their behaviour it was often hard to do this due to work or other unavailability.  

The authors state that their personalized approach did not lead to a higher willingness to use the program than other untailored approaches used in other experiments. The schedules of the users could be taken into account to better approach users for effective savings. If the savings are only very small other incentives should be possibly used such as a widespread reduction in CO2 emissions.

Integration of electric vehicles in smart grid: A review on vehicle to grid technologies and optimization techniques

Bidirectional V2G, vehicle to grid, has many advantages such as reduce power grid losses, prevent power grid overloading, minimize emissions, maximize profits and help intermittent renewable energy. The drawbacks however are battery degradation, the need for more complex hardware, a high investment cost and social barriers, because people want their car to be charged in case of emergency. Next to cost advantages the vehicle to grid also gives the owners backup power in case of a blackout. The article states that due to battery degradation V2G has to either be completely avoided or correctly optimized in order to only discharge the batteries slowly and till at max 60 percent of full capacity. This maximal percentage also works well taking into account social barriers where car owners want a certain battery level at all times. Finally, the authors state that an incentive-based system is needed for V2G to be taken up by many electric vehicle owners.  

The effects of household automation and dynamic electricity pricing on consumers and suppliers

The article states that the amount of savings done by household automation depends on the household's energy consumption and production through the day. It also depends on the size of the household how much savings can be done and if they are done at all. The presence of solar panels can in fact lead to less profits for a single person household. The automation of households leads to savings in both TOU and RTP pricings, with more savings with RTP. Especially profits made with solar panels can be a great incentive for the households. The suppliers however have a decrease in profits due to the increase in solar panels and automation of the households. Thus, suppliers should tailor their contracts to the consumers if the suppliers want to maximize their own profits.  

Forecasting day-ahead electricity prices: A review of state-of-the-art algorithms, best practices and an open-access benchmark

This paper sets out to find the state-of-the-art electricity price forecasting models, it describes problems that make comparing of different models hard and also state that there is no clear benchmark to check the performance of models to. The paper states that there are three main models, statistical models, machine learning models and hybrid models. The comparison of these 3 is very hard thus leading to the authors stating not one single state-of-the-art method but choosing multiple. For the statistical models the authors decide that the LEAR model is very accurate, while for the machine learning models the DNN model is most state-of-the-art. The hybrid models they state to be not compared enough to other models thus they decide to leave them out of consideration.  

LEAR stands for Lasso Estimated Auto Regressive, where Lasso is a regression analysis method that performs both variable selection as well as regularization, which is useful to increase the quality of the dataset used for the model. Auto regressive just points to the type of model being based on time series analysis.  

DNN stand for Deep Neural Network, which is a type of machine learning with the objective of trying to replicate the way a human brain thinks. The deep part stands for it being multiple levels of machine learning. These models can be better in analysis and prediction than the statistical models but do use much more computing power.  

Residential Demand Response Based on Dynamic Electricity Pricing: Theory and Practice

This paper is very long but gives a few good insights into dynamic pricing, it shows that dynamic pricing does indeed reduce the load of mid-peak and peak power plants over the year, while also adding that the peak power plants have to indeed be turned on less due to dynamic pricing. The dynamic pricing also apparently works best when there is already a large amount of renewable and uncontrollable sources of energy.  

Further it states that correct meters are needed in household in order to make sure that the electricity need is measured correctly for every hour such that the dynamic pricing can work well. The addition of online monitoring, graphic user-interfaces and in-house displays can help provide useful information to make dynamic pricing work more efficient. The automation is also a wish of households such that dynamic pricing is correctly triggered, and it does not affect the comfort of people, it also makes the use of dynamic pricing more reliable leading to a better integration of the demand spreading. Finally, the author states that substantial savings can be made but this depends on if an electric vehicle is involved and other factors. A balance has to be found between prices and practicality.

Demand side management of industrial electricity consumption: Promoting the use of renewable energy through real-time pricing

This paper looks at the effectiveness of demand side management of electricity usage, which is thus using real time pricing, but in the case of large industrial electricity consumers. The industrial consumer consisting of a highly efficient cold storage were able to decrease their electricity costs dramatically by looking at the day ahead pricing of the electricity. This thus led to an increase in the amount of wind energy that was used since this corresponded to the electricity prices. Another industrial consumer, a manufacturing plant, showed to not benefit from real time pricing due to a very inflexible load profile, thus leading to much electricity use at high load times.  

The study/paper found a clear relation between the adaptation of consumers to the dynamic pricing and an increase in the use of renewable wind energy, thus showing that an increase of users of dynamic pricing would often increase the consummation of all the renewable energy. The dynamic pricing must however give a financial gain otherwise the consumer is not rewarded enough for the effort needed to follow the electricity demands.

Electric vehicle charging and discharging scheduling strategy based on dynamic electricity price

The paper set out to make a deep learning algorithm capable of most optimally charging an electric vehicle depending on the dynamic electricity prices. It states that the use of this algorithm more evenly spreads the charging of the vehicles to low-peak times while still having the requirement for the vehicles to always be fully charged. The use of the algorithm led to more robustness in the electricity usage since the addition of more vehicles would not lead to a huge increase in peak loads. The algorithm used also led to big savings for the users saving them much money on the charging of their electric vehicles, while also making more use of renewable sources of electricity. The authors state that the algorithm also takes into account the flexibility of charging and discharging. The paper however does not take into account the degradation of the batteries of the electric vehicles, this should be taken into account since this contributes much to the acceptability of the algorithm. In total though the algorithm used would lead to better spread of electricity usage, better integration of renewable energy and would reduce power grip consumption.

|||CITATION NEEDED||| Summary of paper on the effects of dynamic electricity pricing and the use of an app on costs and emissions |||CITATION NEEDED|||

The paper has looked into the extent that an app has effect on making a large group of test households adapt to dynamic electricity pricing. The authors looked at how flexibility of the electricity demand can have an effect on electricity usage in both a case with the use of an app stimulating correct behaviour and without such an app. The use of more elastic electricity usage led to a decrease in total system costs and led to an increase in the renewable energy usage. Although comparing to a fully renewable energy focused setup the elasticity of the groups led to a bit more carbon emissions. The use of the app led to an increase of 0.6 percent in total reduction of the system costs, going from 2 percent to 2.6 percent. The elasticity in the electricity use could in some cases lead to more peaks and less smoothness, these peaks however are not on prime times and would not lead to a large need for more fossil fuel-based electricity sources.  

Analyzing the impact of dynamic electricity prices on the Austrian energy system

The main findings found in the paper on the experiment run with households using an informative app on electricity usage and prices, sending a part of the households a message when the price was lowest, and savings could be most achieved. The research found that a discount message led to an increase in electricity usage of around 1 percent on weekdays and 0.8-2 percent on weekends, where messages stating environmental reasons may elicit more wanted behaviour. Households which used the app heavily had even higher savings of up to 7 percent.  

Peer-to-peer comparison shown in the app led, according to the paper, to a decrease in electricity usage for many households. A gaming functionality which was also added also seemed to contribute to more savings for the households and thus more usage of the available renewable energy. The test also showed that users start out using the app more and after time learn and thus have to use the app less.  

The research state that an app needs: a dynamic tariff scheme, frequent and easily accessible information on marginal prices and frequent and easily accessible information on the household's electricity consumption. Finally, they state that the consideration of ‘smart’ appliances should be taken into account since automation can play an important role in the future.  

My Phone and Me: Understanding People’s Receptivity to Mobile Notifications

A paper from 2021 reporting on a quantitative research using a mobile phone app researches the behaviour of mobile phone users when interacting with notifications. A. Mehrota et al. constructed a moblie app that monitors mobile phone notifications, the way the users interact with these notifications and why they chose to act like they did. They app measured the following: The time it took for the notification to be seen by the user, or the "seen time"; The time between the user seeing the notification and acting on, or the "decision time". The following was asked in a survey, which the users would sporadically receive and fill in: The way the users would respond to these notifications; The reason why they acted like they did; The complexity of the task they were peforming while receiving a message; The effect the notification had on the user; The general personality traits.

The paper found several connections between these measurements and/or surveys, such that it would be too much to mention all of them in this summary. Becuase of that, we will be listing some connection that were found in this paper which are usefull for our own project: Users would see a notification quicker when perfoming a more complex task: Users would experience more disruptiveness from a notification when perfoming a more complex task; Users would open the notification, despite performing a complex task, if the notification contained usefull information; Users would experience more disruptiveness when receiving a notification when they almost finished a task.

This paper concludes some points which might be usefull to keep in mind when designing our own product: If we want to make sure the notifications our app sends to the user are seen quickly, acted upon quickly and not cause disruptiveness for the user, the best time to send our notification is when the user has just started running errands. The first reason for this is the complexity of the task: The task may not be that complex, but it is a relatively complex task a user might perform at home, which is where they should perform the actions suggested in our notification, and the complexity of the task will lead to a shorted time that it will take the user to see the notification. The second reason for this is that a user would experience less disruptiveness when receiving a notification at the beginning of a task. One other thing we will need to keep in mind is that the user is most likely to act on our notification when it contains usefull information, which is why we should make sure to present some information which is important for the user.

Appendix

Appendix A: Planning and logbook

Appendix A.1: Planning

Milestones and responsibilities

	Week 1	Week 2	Week 3	Week 4	Week 5	Week 6	Week 7	Week 8	Week 9
Decide on subject of the course	All
Preparing meeting agendas and make minutes	J
State-of-the-art section	L
RPC section	M
Deliverables section	S
Logbook and planning	J
Study on cost difference between energy contracts		S
Study on the impact of our product on network congestion		J
Study on consumer behaviour		L/M
Ideation of the design of the product		S
Construct the problem statment			L
Interviewing potential users			M
Contact companies involved in network congestion			J
Finalize the algorhythm needed for the product			S
Summaries of literature studies on the wiki				L
Make the final decicion on the programming language used for the product					S
Analyse the interviews of users					J
Analyse the contact of companies					J
Finalize the app for product testing						S
Prepare presentation							J
Discussion							J
Future research							L
Appendix							J
Bibliography							L
Process feedback on presentation and finalize presentation							J
Test the product							M
Final presentation								L
Process the product test								M
Finilazation of the wiki									J

Appendix A.2: Logbook

Week	Name	Break-down of hours	Total hours spent
1	Sven Bendermacher	Searing for ideas (2h), Meeting about subject (1h), Writing deliverables section and mail teachers (0.5h), finding/scanning some promising literature [1-4] (2.5h).	6
	Marijn Bikker	Introductory lecture, research into problems and possible technical solution, Meeting about subject, writing problem statement and RPC's.	6
	Jules van Gisteren	Searching for ideas (1.5h), Preparing meeting (0.5h), Meeting about subject (1h), Creating the logbook and planning (2h)	5
	Lin Wolter	Searching for ideas (2.5h), Looking into possible users (2h), Start of literature study with writing of State-of-the-art (3.5h)	8
2	Sven Bendermacher	Meeting on Monday (2.5h), Looking at possible devices and how to use (2h), Working at the layout and design of the app (4h)	8.5
	Marijn Bikker	Meeting with tutors, working together on problem, literature study, meeting, literature research, user study	7.5
	Jules van Gisteren	Meetings (2.5h), Literature study (3h), Research on possible parties involved in subject (1.5h)	7
	Lin Wolter	Working in group (2h), Finding research on consumer wishes and summarising (4h), Writing problem statement and some other alterations of wiki (1.5h)	7.5
3	Sven Bendermacher	Meetings (3h), Finding energy data on house appliances (2h), Finding information about smart devices connections protocols (2h), Programming a working convolution cost algorithm (3h), starting on learning to code a android app in python (2h).	12
	Marijn Bikker	Discussing the project, discussing the website, finishing interviews, writing informed consent form, meeting, interviewing. (more hours planned next week)	5
	Jules van Gisteren	Meeting on monday (2h), Meeting on thursday+preparations (1.5h), Creating planning (1h), Creating mail to Enexis and processing feedback on the mail (0.5h)	5
	Lin Wolter	Meetings on monday and thursday (3h) Expansion of problem statement (1h) More literature research (1.5h) Finding of data on electricity usage of appliances (2h)	7.5
4	Sven Bendermacher	Meetings on today and working in group (3h), Meeting Thursday (1h), learning app-loading (1h), automating data fetching for the algorithm (2h). (Due to illness I didn't do anything during the weekend)	7
	Marijn Bikker	Meeting on monday and working in group (3h), Writing transcript interview(0.5h), Meeting thursday(1h), learning app-coding(2,5h)	7
	Jules van Gisteren	Meeting on monday and working in group (3h), Meeting thrusday (1h)	4
	Lin Wolter	Meeting on monday and working in group (2h), Doing interview (0.5h), Meeting thursday(1h), Writing scenarios and finding needed information(3.5h)	7
5	Sven Bendermacher	Programming base layer of app (3h), Meeting on Thursday (1h), Programming setting layout (1h), Programming home page (3h).	8
	Marijn Bikker	Meeting on monday, programming, designing.	5.5
	Jules van Gisteren	Meeting on monday, Literature study, Meeting on thrusday, Rewriting requirements, Ordering the document, Rewriting and making additions to Users	8
	Lin Wolter	Meeting on thursday, Literature study and rewriting/ordering of state of the art. (Less meeting due to illness)	7
6	Sven Bendermacher
	Marijn Bikker	Meeting on monday, programming, Meeting on thursday, programming.	6
	Jules van Gisteren	Meeting on monday, processing feedback, partitioning tasks, meeting on thursday and partitioning tasks, fixing the citations on the wiki, reworking the section headings and order	6
	Lin Wolter
7	Sven Bendermacher
	Marijn Bikker
	Jules van Gisteren
	Lin Wolter
8	Sven Bendermacher
	Marijn Bikker
	Jules van Gisteren
	Lin Wolter

Appendix B: User interviews

Appendix B.1: Informed consent interviews users

You have been asked to participate in a study for the course 0LAUK0 Project robots everywhere(2023) of Eindhoven University of Technology. This document gives you information about this study and your rights as a participant. Please read it carefully.

About the study

The aim of this study is to test the need for an algorithm and app using a dynamic electricity contract to steer the user towards low-use hours. The study will last approximately 20 minutes. In this study, you will be asked some questions about your opinion and preferences for such a tool. The experiment leader will make notes of what you are saying.

Voluntary

Your participation is completely voluntary. You can refuse to participate without giving any reasons and you can stop your participation at any time during the study. You can also withdraw your permission to use your data up to 24 hours after the study is finished. All this will have no negative consequences whatsoever.

Confidentiality

All research conducted at the Human-Technology Interaction Group adheres to the Code of Ethics of the NIP (Nederlands Instituut voor Psychologen – Dutch Institute for Psychologists). We will not be sharing personal information about you to anyone outside of the research team. No video recordings are made that could identify you. Only an audio recording will be made. The information that we collect from this study is used for writing scientific publications and will only be reported at group level. It will be completely anonymous and it cannot be traced back to you

Further information

If you want more information about this study you can ask Marijn Bikker (contact email: m.w.a.bikker@student.tue.nl). If you have any complaints about this study, please contact the supervisor, m.j.g.v.d.molengraft@tue.nl.

Certificate of Consent

I, (NAME)……………………………………….. have read and understood this consent form and have been given the opportunity to ask questions. I agree to voluntarily participate in this study carried out by group 1 of the course 0LAUK0 Project robots everywhere(2023).

Participant’s Signature: .........

Date: ...........

Appendix B.2: Explanation before interview

Een dynamisch contract is een vorm van energiecontract waarbij (in het geval van de Nederlandse variant) één dag van tevoren de energieprijzen voor de volgende dag worden vrijgegeven. Deze prijzen verschillen per uur, waarbij de prijzen vaak hoger liggen bij rond piek uren waar mensen over het algemeen veel stroom gebruiken - denk aan de tijd rond het avondeten -, maar waar de prijzen stukke lager - of zelfs negatief zijn! - rond de tijd waarop weinig stroom wordt gebruikt en veel groene energie wordt opgewekt.

https://www.dynamisch-tarief.nl/stroom/ (The site doesn't allow images to be uploaded at this time, so I'm putting this in

Appendix B.3: Interest interview

Interview mensen met een dynamisch contract:

1.      Heeft u het informed consent form begrepen gelezen?

2.     Heeft u apparaten waar je de tijd op kunt instellen voor gebruik? De vaatwasser en wasmachine bijvoorbeeld?

3.     Heeft u een veelgebruiker (elektriciteit) die ‘smart’ is? Dat wil zeggen met wifi verbinding maakt en met een app te bedienen is?

4.     Heeft u een elektrische auto?

5.     Heeft u zonnepanelen?

6.     Wat zou u vinden van een app die automatisch apparaten aanzet op de goedkoopste momenten?  

7.     Wat zou u vinden van een app die elke dag 1 of 2 meldingen stuurt over wanneer de stroom het goedkoopst is?

8.     Heeft u een dynamisch contract, zo ja wat voor soort?

9.     Wat zou u belangrijk vinden aan de app, gebruikersgemak, looks of mogelijkheid tot personalisatie?

10.   Wat zou u vinden van een feature die bijhoud hoeveel geld er dit jaar is bespaard?

Beantwoord de volgende vragen van een schaal van één tot vijf, waarbij één zeer oneens is en vijf zeer mee eens.

11.  “Ik ben overgestapt op een dynamisch contract vanwege financiële overwegingen”

12.  “Ik ben overgestapt op een dynamisch contract vanwege milieu overwegingen”

Interview mensen met een vast of variabel contract:

1.      Heeft u het informed consent form begrepen gelezen en ingevuld?

2.      Heeft u apparaten waar je de tijd op kunt instellen voor gebruik? De vaatwasser en wasmachine bijvoorbeeld?

3.      Heeft u een veelgebruiker(elektriciteit) die ‘smart’ is? Dat wil zeggen met wifi verbinding maakt en met een app te bedienen is?

4.      Heeft u een elektrische auto?

5.      Heeft u zonnepanelen?

6.      Wat zou u vinden van een app die automatisch apparaten aanzet op de goedkoopste momenten?

7.      Zou u een app interessant vinden die elke dag 1 of 2 meldingen stuurt over wanneer de stroom het goedkoopst is?

9.      Wat zou u belangrijk vinden aan de app, gebruikersgemak, looks of mogelijkheid tot personalisatie?

10.   Wat zou u vinden van een feature die bijhoud hoeveel geld er dit jaar is bespaard?

Beantwoord de volgende vragen van een schaal van één tot vijf, waarbij één zeer oneens is en vijf zeer mee eens.

11.     “Als ik zou overstappen op een dynamisch contract, dan zou ik dit doen vanwege financiële overwegingen”

12.     “Als ik zou overstappen op een dynamisch contract, dan zou ik dit doen vanwege milieu overwegingen”

Transscripts

Interview 1

interviewer: Marijn Bikker

Interviewee: Floris Bikker

Interview mensen met een dynamisch contract:

1.      Heeft u het informed consent form begrepen gelezen?

Ja die is gelezen en begrepen.

2.     Heeft u apparaten waar je de tijd op kunt instellen voor gebruik? De vaatwasser en wasmachine bijvoorbeeld?

Ja die hebben we.

3.     Heeft u een “veelgebruiker“ (elektriciteit) die ‘smart’ is? Dat wil zeggen met wifi verbinding maakt en met een app te bedienen is?

Nee dat hebben we niet.

4.     Heeft u een elektrische auto?

Nee

5.     Heeft u zonnepanelen?

Nee

6.     Wat zou u vinden van een app die automatisch apparaten aanzet op de goedkoopste momenten?

Handig, maar het zou een probleem kunnen zijn op de momenten wanneer de apparaten niet klaar zijn voor gebruik. En bovendien als je al bezig bent met de apparaten, dan is het nog maar een kleine moeite om zelf ook de tijd in te stellen wanneer hij moet draaien.

Voor elektrische auto echt handig.

Nu is het zelf in de energie-app kijken een klein beetje gedoe en gereken, een app die dat zelf doet zou misschien makkelijk zijn. Nu is het nog nieuw en leuk om in die app het even op te zoeken en uit te rekenen, misschien dat dat over een tijd minder leuk om te doen is.

7.     Wat zou u vinden van een app die elke dag 1 of 2 meldingen stuurt over wanneer de stroom het goedkoopst is?

Best handig.

8.     Heeft u een dynamisch contract, zo ja wat voor soort?

Ja, per uur variërende prijs.

9.     Waarom bent u overgestapt op een dynamisch contract, voor besparing van geld, het milieu of een andere reden?

Beiden wel. Geld en milieu. Tijdens piek uren wordt er natuurlijk elektriciteit opgewekt met fossiele brandstoffen, als wij dan elektriciteit kunnen gebruiken tijdens uren waarop er meer groene stroom is dan scheelt dat voor het milieu.

10.  Wat zou u belangrijk vinden aan de app, gebruikersgemak, looks of mogelijkheid tot personalisatie?

Gebruikersgemak het belangrijkst. Hoe de app eruit ziet niet zo.

11.  Wat zou u vinden van een feature die bijhoud hoeveel geld er dit jaar is bespaard?

Ja leuk.

Research on profit of dynamic contract

To get costumers to switch to a dynamic contract, it is important to study the possible profit of switching to such a contract.

We look at a period of at least a year, since the price one pays in the dynamic case differs a lot in the summer and winter because of the difference in electricity consumption. To compare a fixed contract with a dynamic contract, some assumptions have to be made. The most important assumption is that the market price of the electricity stays more or less the same throughout the years. The market price can change due to global (in)stability or big events, such as wars or pandemics affecting the financial market. For our comparison we assume we are dealing with a stable market price.

The business works as follows. The energy suppliers sell the electricity to the customers with a fixed price, whilst they buy the electricity on the market for fluctuating prices. In general the price the costumer pays is higher than the price the energy suppliers pay, since the energy suppliers have to make profit. But for the fixed or variable contracts the energy suppliers work with an additional safety margin for the case of rising prices. In the case of the dynamic contract there is only the small addition in price for the energy suppliers to make profit.

Then there is a risk involved in switching to a dynamic contract. You have to pay more when the market prices increase, whereas someone with a fixed contract is not affected by this, until the moment his contract expires and he has to sign a new contract.

The costumer with the dynamic contract can thus profit from the moments the market price is lower than the price the fixed contracts offer. To profit from this, the costumer must be active and use these moments to use electric devices or load their laptops and phones. The study of the European commission gives insight into the possible costumers that use a dynamic contract profitably.

Old interview that wasn't conducted: Usability requirements interview

Two requirements of our product are based on the experience of the user of the product:

• The user should not feel disrupted by the app;
• The environment in which the user interacts with the product should be easy to navigate, such that users with every level of experience with technology should be able to know what they are telling the appand what the app is telling them.

Becuase we cannot easily check if our product contains these requirements - we as creators might not feel disrupted by the product and we already have a good idea of how the app works becuase we designed it - we are going to check for these requirements using an interview. Since we already want to gain some insight into this requirement before we finished a working version of the product, we are going to conduct this interview on the hypothetical use of this product. Before this interview, the interviewees will have signed the informed consent form and have been told some preliminary knowledge about the state-of-the-art of our product, which can be found in Appendix B.1 and Appendix B.2 respectively.

The following questions should be answerd using the Likert scale (i.e. choosing between strongly disagree, disagree, neutral, agree or strongly agree). The questions are - as of now - written in Dutch. In the future we will include both an English and Dutch version.

1. Een app die mij verteld wanneer het beste moment zou zijn om geld te besparen met een dynamisch contract zou mij eerder over laten stappen tot een dynamisch contract.
2. Ik heb liever dan een algoritme checkt wat het beste moment is om mijn apparaten aan te zetten dan dat ik dat zelf zou doen.
3. Een app die op een vast moment op de dag een notificatie stuurt op mijn telefoon wanneer ik het beste mijn apparaten (die stroom gebruiken) (vaatwasser, wasmachine, droger, etc.) aan kan zetten om geld te besparen zou ik prettig vinden.
4. Een app die mij op een scherm in een app kan laten zien wanneer ik het beste mijn apparaten aan kan zetten zou ik prettig vinden.
5. Ik zou liever een notificatie krijgen die mij herrinert wanneer ik mijn apparaten aan moet zetten dan dat ik zelf op een scherm moet checken.
6. Ik zou een notificatie op een door mij zelf ingesteld op de dag kunnen instellen zodat deze altijd