Energy System Operation
In CITIES, we have developed controllers and operational strategies for direct control of system states like temperature control in district heating for short-term/operational use.
The controllers are based on probabilistic forecasts and combined with operational strategies, for example, price signals to move end-users’ energy consumption and relieve the energy system through flexibility.
Findings in the CITIES project – possible benefits (CO2 + Costs) from using digitalization in the operation system:
- Up to 800 mill Dkr annually savings in Denmark by data-driven temperature optimization (and tons of CO2 savings) in the district heating system. See the report here
- 10-30 pct savings by predictive control of heat pumps
- 5-15 pct savings by integrating forecasts in smart house controllers
- 10-40 pct improvements in electricity and heat load forecasts
- Up to 10 pct savings by optimal operations of CHP and DH plants
- 5-12 pct savings by smart wastewater treatment
- 10-20 pct savings on cooling for supermarkets
- Up to 90 pct savings on cooling for data centres
The CITIES partners have defined six recommendations for this topic: Digitalization and data-driven operation of integrated energy systems are key to the green transition; Spatial-temporal thinking and coherency is important; We must prioritize digital and automatic solutions; System solutions for sector coupling and Power2X (incl. Power2Heat) should be focussed and barriers for those solutions must be eliminated; We must consider using virtual storage solutions; National initiatives for digital and cross-sectorial solutions should be supported and expanded.
Learn more in the list below and see the videos..
Flexibility / virtual storage – a keyword
Flexibility / virtual storage is a keyword to green transition in a cheap and effective way and in planning and operation.
To obtain the Danish climate goals for 2030 and 2050 we need to focus on digitalization and sector coupling. Digitization is the link that enables the coupling of energy sources with real-time matching of energy demands and production through data intelligence and IoT devices in integrated energy systems.
In the CITIEs project, we have investigated how much flexibility can help us achieve climate goals.
We note these characteristics for flexibility:
- Flexibility (or virtual storage) characteristics:
- Wastewater systems can provide storage 0.2-6 hours ahead
- Supermarket refrigeration can provide storage 0.5-2 hours ahead
- Buildings thermal capacity can provide storage up to, say, 2-10 hours ahead
- Buildings with local water storage can provide storage up to, say, 2-18 hours ahead
- District heating/cooling systems can provide storage up to 1-3 days ahead
- DH systems can provide seasonal/long term storage solutions
- Gas systems can provide seasonal/long term storage solutions
https://www.youtube.com/watch?v=hCwOUGggyvw&feature=youtu.be
Model Predictive Control for Smart Energy Systems
CITIES has developed tools for short term (probabilistic) forecasting and control of integrated energy systems with flexible geographic scope. At CITIES final conference Professor John Bagterp Jørgensen talked about the software solutions for predictive control (MPC). The technology is based on stochastic differential equations (SDEs)
https://youtu.be/rVwx8FVSxUA
Professor Henrik Madsen: Forecasting for the Green Transition
In future energy forecasting will play a central role in the green transition, when demand follows the renewable power production. In this talk Project Manager at CITIES, Professor Henrik Madsen, describes state-of-the-art methodologies for renewable energy forecasting. He also describes how integrated forecasting across domains (wind, solar, load, prices, …) will become essential.
https://youtu.be/UpGns2XWIgQ
Characterisation and Integration of Energy Flexibility
Flexibility in electricity consumption will be a keyword for the green transition, in the energy system of the future based on solar and wind. Postdoc Rune Grønborg Junker from DTU Compute talks about how to unlocking consumers flexibility potential.
https://www.youtube.com/watch?v=OeZx1skvW3A&feature=youtu.be
Demand response solutions and perspectives
CITIES Final conference November 9, 2020: Demand response solutions and perspectives Ben Kroposki, Director, National Renewable Energy Lab, Colorado, US
https://www.youtube.com/watch?v=H88eCFSwCto&feature=youtu.be
Smart Cooling – Singapore, Grindsted and the Future
At CITIES final conference November 9, 2020 postdoc Dominik Franjo Dominkovic talked about the research we have done in Singapore to utilized excess waste heat via absorption shillers to generate cold and analyzed the potential of widely adopted district cooling in tropical regions, with steady cooling demand. He has used the same methods on a demo case in Grindsted for cooling meeting rooms and the research continues now in ‘Cool-Data’ – with data centres.
https://www.youtube.com/watch?v=6qHZ9_4FWso&feature=youtu.be
Flexibility concepts in practice; case study on water towers
Rune Grønborg Junker, Postdoc at DTU Compute, together with colleagues and with the help of Grundfos, has found an intuitive way in which mathematical models describe how energy-flexible consumers react to electricity prices. They have used a water tower at Grundfos to investigate price-based management of flexibility.
https://www.youtube.com/watch?v=UyVc7w9zq5s&feature=emb_title
Using SDEs to understand energy efficiency in buildings
Jaume Palmer Real is a PhD student at DTU COMPUTE – the Technical University of Denmark. In his YouTube-video, he tells about his PhD project: Stochastic Differential Equations (SDE) for Modeling Energy Systems Integration – and about how to use SDEs to understand energy efficiency in buildings. A model we use in CITIES.
https://www.youtube.com/watch?v=X8BTz91xmz8&feature=youtu.be
Smart Control of Wastewater Treatment Aeration
In a CITIES demo project, Krüger-Veolia together with DTU Compute has tested how to use the wastewater plant to increase flexibility in the energy system and to save CO2 and money. At the Green Digitalization conference on November 10, 2020, Peter Stentoft Krüger talked about the results.
https://www.youtube.com/watch?v=dAdNDneyWX8&feature=youtu.be
Data-driven Methods to Characterize the Dynamic of Buildings
At CITIES final conference on November 9, 2020, Carsten Rode, Professor at DTU Byg and Work Package Manager on CITIES & Christoffer Rasmussen, PhD at DTU Compute explained how they have developed a concept whereby they use mathematics/statistics to analyze data collected with high frequency from sensors in the buildings to optimize energy efficiency and to use the buildings as energy flexible storage.
https://www.youtube.com/watch?v=SGihJZy9xZs&feature=youtu.be
Methods for Optimal Operation and Market Participation of DH
District heating provider with a portfolio of production units including combined heat and power (CHP) plant. A novel bidding method (HURB) has been developed to optimize the daily production of heat to cover the heat demand at minimal cost and to sell the electricity from the CHP, if beneficial for the overall system cost. At CITIES final conference Daniela Guericke (DTU) & Anders Andersen (EMD International) talked about the research.
https://www.youtube.com/watch?v=aSYgtMv_2Hw&feature=youtu.be
Impact of Energy Communities on Distribution Grids
In a CITIES demo project, Dansk Energy together with DTU Computes have made a simple analysis of the impact of energy
communities on three different grid layouts to estimate the consequences of different setups of energy communities on the distribution grid. At the Green Digitalization conference on November 10, 2020, Tilman Weckesser, Dansk Energi, presented the results.
https://www.youtube.com/watch?v=6EMhkxFmvZ4&feature=youtu.be
Leveraging consumers’ flexibility for the provision of ancillary services
Flexibility in electricity consumption will be a keyword for the green transition, in the energy system of the future based on solar and wind. Giulia de Zotti from the Danish Energy Agency and DTU Compute talks here about how to unlocking consumers flexibility potential.
https://youtu.be/RZi5RjDKAH0
Videos: Forecasting and Control – Towards Digital Grids: Lesson 1: Introduction to renewable energy forecasting, (MET
input, data, adaptivety, combined forecasting, etc.)
Get updated on the state of the art methods for Energy Systems Integration in these videos with Professor Henrik Madsen from DTU Compute – the section for Dynamical Systems. The videos are made in collaboration with EIT InnoEnergy at the university KU Leuven in Belgium.
Get an introduction to forecasting methods – and learn about digitization of the grids and energy markets. Find the references here
https://youtu.be/0IiVgpELtjM
Videos: Forecasting and Control – Towards Digital Grids: Lesson 2: Point forecasts of wind and solar power production
Get updated on the state of the art methods for Energy Systems Integration in these videos with Professor Henrik Madsen from DTU Compute – the section for Dynamical Systems. The videos are made in collaboration with EIT InnoEnergy at the university KU Leuven in Belgium.
Get an introduction to forecasting methods – and learn about digitization of the grids and energy markets. Find the references here
https://youtu.be/Z-E_TDLIx8c
Videos: Forecasting and Control – Towards Digital Grids: Lesson 3: Probabilistic and full stochastic forecasting
Get updated on the state of the art methods for Energy Systems Integration in these videos with Professor Henrik Madsen from DTU Compute – the section for Dynamical Systems. The videos are made in collaboration with EIT InnoEnergy at the university KU Leuven in Belgium.
Get an introduction to forecasting methods – and learn about digitization of the grids and energy markets. Find the references here
https://youtu.be/Pjyv49LaKMc
Videos: Forecasting and Control – Towards Digital Grids: Lesson 4: The challenges of the climate crisis and the integration of renewables
Get updated on the state of the art methods for Energy Systems Integration in these videos with Professor Henrik Madsen from DTU Compute – the section for Dynamical Systems. The videos are made in collaboration with EIT InnoEnergy at the university KU Leuven in Belgium.
Get an introduction to forecasting methods – and learn about digitization of the grids and energy markets. Find the references here
https://youtu.be/vn94ODtst5U
Videos: Forecasting and Control – Towards Digital Grids: Lesson 5: Unlocking end-user flexibility, (description of flexibility, use of flexibility for, demand response, etc.)
Get updated on the state of the art methods for Energy Systems Integration in these videos with Professor Henrik Madsen from DTU Compute – the section for Dynamical Systems. The videos are made in collaboration with EIT InnoEnergy at the university KU Leuven in Belgium.
Get an introduction to forecasting methods – and learn about digitization of the grids and energy markets. Find the references here
https://youtu.be/AR5F77dTgJw
Videos: Forecasting and Control – Towards Digital Grids: Lesson 6: Data-intelligent operation of future smart energy systems
Get updated on the state of the art methods for Energy Systems Integration in these videos with Professor Henrik Madsen from DTU Compute – the section for Dynamical Systems. The videos are made in collaboration with EIT InnoEnergy at the university KU Leuven in Belgium.
Get an introduction to forecasting methods – and learn about digitization of the grids and energy markets. Find the references here
Webinar: Accelerating the green transition using AI and energy system integration
The green transition calls for the next level of digitization and a need for new tailored methods, i.e. energy system-oriented AI, Big Data Analytics, Grey-Box Modelling, Cloud-Fog-Edge Computing, System-of-Systems, IoT, Resilience, User-involvement, and Data-sharing principles. This special CITIES session provides a number of talks highlighting the status and challenges in achieving an efficient implementation of the future low-carbon society.
Webinar: Data-Driven Technologies for Energy Efficiency and Flexibility
Contemporary digital and data-driven methods are key to optimize the operation of buildings and district energy systems such that use of fossil fuels can be minimized without compromising on comfort and functionality of the built environment. The CITIES workshop gives an overview of potentials and state-of-the-art technologies in the field.
Recommendations
-
Digitalization and data-driven operation of integrated energy systems are key to the green transition
Digitalization and data-driven solutions should be prioritized at all aggregation levels of the energy system in order to ensure an efficient and secure operation of the future weather driven low-carbon energy system. Data-driven solutions are needed for providing an efficient integration of the energy systems (power, heat, gas) as well as for the interaction with the water and food sectors. -
Spatial-temporal thinking and coherency is important
It is important to adopt a spatial-temporal systems of systems framework when considering solutions for the future weather-driven energy system. In CITIES we have suggested the Smart-Energy Operating-System (SE-OS) for providing this. The coherence of forecasts, aggregation, models, etc., between all spatial-temporal scales is crucial. -
We must prioritize digital and automatic solutions
We have demonstrated a large number of opportunities to unlock flexibility and hence to reduce CO2 drastically by using digital and automatic solutions. Therefore, digitalization can accelerate the green transition through implementation of green digital solutions. -
System solutions for sector coupling and Power2X (incl. Power2Heat) should be focussed and barriers for those solutions must be eliminated
Research results clearly show that the integration of different energy sectors facilitates the flexibility in the energy system which is crucial for the integration of intermittent renewable energy production from e.g. wind and solar. One example is that excess wind power can be stored as heat in DH systems. -
We must consider using virtual storage solutions
Integrated and digital operation can lead to (virtual) storage solutions on all relevant time scales from minutes, days to seasons. We can store energy in the thermal mass of buildings, or switch between biogas and power in a dual heat pump in order to use existing gas storages to ‘save’ green power. Often such virtual storage solutions can be operated without energy losses. -
National initiatives for digital and cross-sectorial solutions should be supported and expanded
It is essential that cross-sectorial solutions are tested in labs, in living labs and in the field and results are consolidated across projects. It is important that tests are representative and scalable. Therefore, organizations such as Uni-Lab.dk, the national umbrella organisation for Living Labs and Test Labs in Denmark, and Center Denmark, the European and National hub for smart and integrated energy systems, should be consolidated and developed further.
Solutions/methodologies
-
Tool for simulating occupancy profiles for private households
ProccS.org is a tool for simulating occupancy profiles for private households. The application creates occupancy profiles for bathroom, kitchen, living room and bedrooms. The profiles can be downloaded in csv-format. -
Software for Model Predictive ControlHPMPC is a toolbox for High-Performance implementation of solvers for Model Predictive Control (MPC). It contains routines for fast solution of MPC and MHE (Moving Horizon Estimation) problems on embedded hardware.
The software is available on GitHub.
MPCR is a toolbox for building Model Predictive Controllers written in R, the free statistical software. It contains several examples for different MPC problems and interfaces to opensource solvers in R . The software is available on GitHub. -
Model for synthesizing data about demand responseDTU Compute and CITIES have developed a model that allows synthesizing data about demand response for a variegate pool of rational electricity consumers (which include commercial, industrial and residential). The model focuses on the scheduling of electricity consumption for different types of rational electrical consumers, achieving cost minimization for consumers while respecting operational and comfort constraints set by consumers. The model has been developed by assuming that rational consumers are equipped with home energy management systems, which receive dynamic electricity prices.
The model is written in GAMS-Matlab and is available under request (please contact gizo@dtu.dk). Moreover, additional documentation is currently under development in version 2, to facilitate the user in the model adoption. The documentation will be uploaded to the CITIES website.
-
Smart-Energy Operation SystemCITIES has developed a Smart-Energy Operation-System SE-OS. The SE-OS is characterised by the indirect control approach based on one-way communication of prices to the end-users and gives a more elaborate description of the flexibility using the so-called Flexibility Function. An aggregator coordinates and operates the production and consumption of its distributed energy resources (DERs) either directly or indirectly eg. by broadcasting a real time price. Contact John Bagterp Jørgensen (jbjo@dtu.dk)
-
Heat Unit Replacement Bidding MethodDistrict heating provider with a portfolio of production units including combined heat and power (CHP) plant. A novel bidding method (HURB) has been developed to optimize the daily production of heat to cover the heat demand at minimal cost and to sell the electricity from the CHP, if beneficial for the overall system cost.
The model HURB first optimizes heat production without market participation by use mixed-integer linear programming and then replaces iteratively heat-only units by CHP production (in descending order of operational costs.
-
Frigg – plug-and-play modelling tool for flexible energy systemsWith the transition towards renewable energy comes the need for flexible energy demand. Frigg is a plug-and-play modelling tool for flexible energy systems. With Frigg, demand side flexibility can be modelled in a way that is interoperable with major energy system modelling frameworks such as TIMES, Balmorel or Calliope.
Frigg uses data from these models, generates hourly prices and simulates the demand-side. Passing an altered demand level back to the energy system model allows computing energy system equilibria that take demand-side flexibility into account in a realistic way. Frigg is currently in early-stage development. The development is financed by the projects CITIES, openENTRANCE and Cool-Data.
Demo projects
Performance monitoring of renewable production assets
In the demo case, CITIES ENFOR and Ørsted have worked with performance monitoring of renewable production assets, considering wind farms as an example.
Qualitative investigation of the impact of energy communities on distr. grids
The demo project has made a simple analysis of the impact of energy communities on three different grid layouts: urban, suburban and rural areas of Denmark to estimate consequences of different setups of energy communities on distribution grids.
Control of heat pumps
The demonstration project has implemented, tested and demonstrated a fluidborne heating system with a heat pump and a solar collector in a domestic building that makes optimal use of renewable energy from the grid.
Thermal mass for energy storage: Impacts and perspectives on system scale
In this demo case, the CITIES project wants to assess or activate/utilize the thermal energy storage potential of different building archetypes, calculate the potential of thermal mass as a storage component on the system scale and assess the impact of utilizing this storage capacity in the energy system as a whole.
Optimization under uncertainty heat/power production in DH systems
This demo case has developed an optimisation method for the heat and power production planning in district heating systems to apply it to a real-world demo-case.
Dynamic prices for heat delivered to district heating systems
This demo project has studied the dynamic value of heat supplied to the district heating system. Many of the methods used and the obtained results could also be used in a study on the demand side.
Smart Meter Data Analytics
The demo case has investigated different approaches to analysis of smart meter data in a way to identify different consumption patterns. The demo case used data from Syd Energi, which had data from about 270.000 smart meters, with 15 minutes to 1-hour resolution and recorded for approximately one year. The idea was that the classification and characterization of electricity consumption should have to give a more detailed picture of consumption patterns to use for estimation and dimensioning of the energy grid.
Smart Energy Systems: Flexible Cooling of Data Centers
Energy demand for cooling of data centres has been increasing steadily in Denmark and the world. On the other hand, the aim of the energy transition is to make the system more efficient, including end-user energy efficiency. In order to continue the energy transition, as well as to integrate a rising number of data centres across Denmark, this demo project will look into possibilities for flexible and intelligent cooling of data centres.
Smart water management to improve water-energy nexus for water supply systems
In the demo case the CITIES project wants to allow water utilities to holistically manage their processes in a flexible way by operational management recommendations focused on: decrease the non-revenue water, improve the pressure management, detect leakages, save energy and water (water-energy nexus), and above all, improve the integrated water resources management in terms of water resources and revenues supporting financial, social and environmental sustainability, while maintaining quality and level of service.
Load forecasting for District Heating (and greenhouses)
By collecting data from Greenhouse consumers and weather observations the demo case has derived an offline model structure for load forecasting in greenhouses and established an online setup for load forecasting in greenhouses. In the online setup, the model parameters have been adapted to the actual dynamics.
Dynamic CO2-based control of summerhouse swimming pool heating
In this demo case CITIES has shown that CO2 emissions can be reduced by at least 10 per cent by heating the water in the indoor swimming pools when the electricity is generated by wind turbines and stopping it when the electricity comes from sources with a higher carbon footprint. This will also make itself felt on the electricity bill, which can usually be cut by EUR 1,400-2,000 (DKK 10-15,000).
Regulating Power Market: Modelling and Forecasting
The demo case has formulated stochastic models for the regulating power market in the Nordic area, however with a special focus on the prices in DK1 and DK2. Ultimately the purpose was to obtain better methodologies for predicting the probabilities for up-or down-regulation, and the related prices and/or volume. Such forecasts have also been used in some other demo projects in CITIES.
Optimal Control of District Heating Supply Temperatures to Greenhouses
The demo case has used research from a parallel demo case on load forecasting in greenhouses, and another demo project on dynamic prices for DH systems to develop methods for optimal control of district heating supply temperature to greenhouses. on load forecasting in greenhouses, and another demo project on dynamic prices for DH systems.
Brochures
Smart Energy Operating System SE-SO
Model Predictive Control in Urban Systems Optimal and Stochastic Utilization of the infrastructure of sewer systems
Performance monitoring of renewable production assets
Dynamic CO2-based control
Data-driven Nonlinear Prediction Model for Price Signals in Demand Response Programs (Lærkevej)
District Cooling – potential of district cooling in hot and humid climates
Characterizing the energy flexibility of buildings and districts
Biomass supply planning for large-scale combined heat and power producers
A framework for prediction and segmentation of daily energy load profiles of building clusters using machine learning
Operational planning and bidding for district heating systems
Smart Control with Embedded Weather Forecasting
Time Constants and Clustering of Building
Leveraging consumers’ flexibility for the provision of ancillary services
Methodology for simple exchange of data between energy operators
