OPTIMUS DSS

Overall Procedure

Contents OPTIMUS DSS GOAL How the OPTIMUS DSS works OPTIMUS DSS setup DSS graphical user interfaces Actions plans

OPTIMUS DSS GOAL The goal of the OPTIMUS project is to help local authorities to optimise the energy performance of public buildings by applying the short-term actions suggested by a Decision Support System (DSS) which handles data obtained in a diversity of sources and domains: Weather conditions Social behaviour Building energy performance Energy prices Renewable energy production

De-centralized sensor (BEMS) How the OPTIMUS DSS works (1/8) Available data Weather forecast De-centralized sensor (BEMS) Occupants feedback Energy prices RES production Data is captured from the buildings and their context. Semantic framework integrates the different data sources using semantic web technologies.

De-centralized sensor (BEMS) How the OPTIMUS DSS works (2/8) Available data Historical data … Sunday Monday Tuesday Wednesday Thursday Friday Saturday Weather forecast De-centralized sensor (BEMS) Occupants feedback Prediction models Energy prices RES production Prediction models use historical data to forecast the building behaviour for the following 7 days.

De-centralized sensor (BEMS) How the OPTIMUS DSS works (3/8) Available data Historical data … Sunday Monday Tuesday Wednesday Thursday Friday Saturday Weather forecast De-centralized sensor (BEMS) Occupants feedback Prediction models Inference rules Energy prices RES production Inference rules use the predicted and monitored data to suggest short-term actions plans to the final user.

De-centralized sensor (BEMS) How the OPTIMUS DSS works (4/8) Available data Historical data … Sunday Monday Tuesday Wednesday Thursday Friday Saturday Weather forecast Raise set point temperature Shift loads at 11 am Partial free cooling at 16 am Start heating system at 7 am De-centralized sensor (BEMS) Occupants feedback Prediction models Energy Models Inference rules Energy prices RES production Short-terms actions plans are presented to the user in a simple and clear manner.

OPTIMUS DSS INTERFACES Action plans suggested by the DSS How the OPTIMUS DSS works (5/8) Available data Historical data … Sunday Monday Tuesday Wednesday Thursday Friday Saturday Weather forecast Raise set point temperature Shift loads at 11 am Partial free cooling at 16 am Start heating system at 7 am De-centralized sensor (BEMS) Occupants feedback OPTIMUS DSS INTERFACES Prediction models Action plans suggested by the DSS Inference rules Energy prices RES production End-users interfaces display the monitored, forecasted data and the short-term plans in order to support experts’ decisions.

OPTIMUS DSS INTERFACES Action plans suggested by the DSS How the OPTIMUS DSS works (6/8) Available data Historical data … Sunday Monday Tuesday Wednesday Thursday Friday Saturday Weather forecast Raise set point temperature Shift loads at 11 am Partial free cooling at 16 am Start heating system at 7 am De-centralized sensor (BEMS) Occupants feedback OPTIMUS DSS INTERFACES Prediction models Action plans suggested by the DSS Inference rules Energy prices RES production The results of the implementation of the actions in each pilot city will modify the data sources

De-centralized sensor-based Feedback from occupants How the OPTIMUS DSS works (7/8) Four different prediction models have been developed and operate within the OPTIMUS DSS to implement the Inference Rules and suggest Action Plans. Weather forecasting RES production De-centralized sensor-based Models implementation: - Multiple linear regressors - Resistance-capacitance model Energy prices Feedback from occupants For each building, an individual configuration is required in order to boost the forecasting performance.

How the OPTIMUS DSS works (8/8) Occupancy AP 1 Scheduling and management of the occupancy AP 2 Scheduling the set-point temperature AP 3 Scheduling the ON/OFF of the heating system AP 4 Management of the air side economizer AP 5 Scheduling the photovoltaic (PV) maintenance AP 6 Scheduling the sale/consumption of the electricity produced through the PV system AP 7 Scheduling the operation of heating and electricity systems towards energy cost optimization Heating and cooling Air conditioning Generation and on-site renewable production

OPTIMUS DSS Setup (1/5) Buildings DSS server Data capturing modules Monitoring data Action plans DSS Monitoring data Action plans - Building Energy Management System DSS server - Monitoring system - Climate conditions - Energy market - Renewable energy production system - Occupants feedback

OPTIMUS DSS Setup (2/5) 1) Registering a new building with some static data

OPTIMUS DSS Setup (3/5) 2) Setting up the building partitioning The zones is a logic model of the reality of the building. Each zone refers to an area of the building which is monitored (through sensors) and/or an area where an action plan can be applied.

OPTIMUS DSS Setup (4/5) 3) Setting up the sensors Name URL service: Web service URL (RapidAnalytics) of the Prediction model used to forecast data. URL: internal identifier of the sensor, Prediction model parameters: List of sensors needed Units Aggregation method

OPTIMUS DSS Setup (5/5) 4 Setting up the action plans The action plans can be invoked for a particular zone (previously defined in step 2). For each zone where the action plan will be applied, the sensors needed by the action plan have to be mapped to the available sensors of the building (Previously defined in step 3).

DSS graphical user interfaces (1/10) End-user environment: Tracker The user’s target for each DSS indicator The potential for improvement The improvement so far

DSS graphical user interfaces (2/10) End-user environment: City Dashboard Mouse hover to see detailed values The city’s DSS indicators for the last week

DSS graphical user interfaces (3/10) End-user environment: Buildings The buildings currently registered on the DSS and the four indicators

DSS graphical user interfaces (4/10) End-user environment: Building Dashboard Mouse hover to see detailed values The building’s DSS indicators for the last week

Unknown action plan status DSS graphical user interfaces (5/10) End-user environment: Action Plans The active action plans for the selected building. Click to see each action plan’s page Unknown action plan status Action plan declined Action plan accepted

Choose the week to be displayed DSS graphical user interfaces (6/10) End-user environment: Historical Data Choose the week to be displayed Click to view The building’s DSS indicators for the selected week Plots with the historical or predicted data of the building’s streams

DSS graphical user interfaces (7/10) End-user environment: Weekly Report Click to edit The building’s weekly reports

DSS graphical user interfaces (8/10) End-user environment: Weekly Report User input for the DSS implementation.

DSS graphical user interfaces (9/10) End-user environment: Weekly Report The action plan status for each day of the week User input for the action plan implementation

DSS graphical user interfaces (10/10) End-user environment: User Activity The recent activity of the DSS users

Action Plans (1/29) Action Plan 1: Scheduling and management of the occupancy General purpose Reduction of the building energy consumption by changing the location of building occupants, so as to use the minimum number of thermal zones and turn off the heating/cooling system in the empty zones. How does it work? By displacing the building occupants to occupy firstly the building zones with the minor energy consumption and secondarily the zones characterized by higher thermal comfort. Indicators to be optimized Energy consumption CO2 emissions Thermal comfort

Action Plans (2/29) Action Plan 1: Scheduling and management of the occupancy Structure of the Action Plan Occupancy: occupation intensity, presence time, constraints related to occupancy Thermal needs Outdoor air temperature Total solar radiation Indoor air temperature Energy consumption Prediction model Historical data Static data Energy consumption Predicted data DSS Application Energy Model Theoretical Inference rules Zaanstad

Action Plans (3/29) Action Plan 1: Scheduling and management of the occupancy Proposed number of occupants Proposed set-point temperature DSS interface Conditioned rooms Building zones “Yes” if the zone has to be open unconditioned rooms

Action Plans (4/29) Action Plan 2: Scheduling the set-point temperature General purpose Optimizing energy use for heating and cooling, while maintaining comfort levels in accepted ranges. How does it work? By supporting the energy manager in adjusting the temperature set-point after taking into consideration thermal comfort parameters. The preferred temperature is calculated through the Thermal Comfort Validation (TCV) and/or the Adaptive Comfort Concept. Indicators to be optimized Energy consumption CO2 emissions Energy cost

Calculation of the Predicted Mean Vote (PMV) set point temperature… Action Plans (5/29) Action Plan 2: Scheduling the set-point temperature Indoor conditions Analysis and evaluation of user’s feedback Set points Calculation of the Predicted Mean Vote (PMV) Reconsider set point temperature… Predicted Values Actual Values Thermal Sensation Building’s users Deviation Inference Rules http://validator.optimus-smartcity.eu ISO 7730:2006 & “A framework for integrating User Experience in Action Plan Evaluation through Social Media”. Proceedings of the 6th International Conference on Information, Intelligence, Systems and Applications (IISA 2015), July 6-8, 2015 - Corfu, Greece.

Action Plans (6/29) Action Plan 2: Scheduling the set-point temperature Historical data Structure of the Action Plan Feedback from occupants Outdoor air temperature Indoor set point temperature DSS Application Energy Model Predicted data Theoretical Inference rules Sant Cugat Savona Zaanstad

Set point temperature suggestion Building zones. Click for details Action Plans (7/29) Action Plan 2: Scheduling the set-point temperature DSS interface Set point temperature suggestion Building zones. Click for details

Action Plans (8/29) Action Plan 2: Scheduling the set-point temperature DSS interface Selected rule and final suggestion TCV rule suggestion Click to see the user’s feedback Adaptive Comfort rule suggestion

Hourly feedback from users of the building Action Plans (9/29) Action Plan 2: Scheduling the set-point temperature DSS interface Hourly feedback from users of the building The colour code used

Action Plans (10/29) Action Plan 3: Scheduling the ON/OFF of the heating system General purpose Reduction of energy use by optimizing the boost time of the heating/cooling system. How does it work? The boost heating phase duration is calculated based on climatic data forecasts and the occupancy of the building. The social feedback and the thermal comfort level is also considered. Indicators to be optimized Energy consumption CO2 emissions Energy cost

Action Plans (11/29) Action Plan 3: Scheduling the ON/OFF of the heating system Structure of the Action Plan On/off of the heating/cooling system Occupied/unoccupied space Space heating capacity Outdoor air temperature Indoor air temperature Energy consumption Social media/ mining Prediction model Static data Indoor air temperature Historical data DSS Application Predicted data Energy Model Sant Cugat Theoretical Inference rules Savona Zaanstad

Action Plans (12/29) Action Plan 3: Scheduling the ON/OFF of the heating system DSS interface Building zones Start and stop schedule of the heating system Proposed set-point temperature 7:00 18:00 7:00 18:00 8:00 18:00 8:00 18:00 7:00 18:00 6:30 18:00 6:30 18:00 6:30 18:00 6:30 18:00 6:30 18:00 7:00 18:00 7:00 17:00 8:00 18:00 8:00 17:00 7:00 18:00 6:00 18:00 6:00 17:00 6:00 18:00 6:00 17:00 6:00 18:00 7:00 18:00 7:00 18:00 7:00 18:00 7:00 18:00 7:00 18:00

Action Plans (13/29) Action Plan 4: Management of air side economizer General purpose Optimizing energy use and reducing energy cost by exploiting outdoor-air to reduce or eliminate the need for mechanical cooling. How does it work? When there is a need for cooling and if the outdoor-air conditions are favorable, outdoor-air is used to meet all of the cooling energy needs or supplement mechanical cooling. Indicators to be optimized Energy consumption CO2 emissions Energy cost

Action Plans (14/29) Action Plan 4: Management of air side economizer Structure of the Action Plan Outdoor air temperature Outdoor relative humidity Indoor air temperature Indoor elative humidity Energy prices Historical data Predicted data Energy Model Theoretical Inference rules DSS Application Sant Cugat

Action Plans (15/29) Action Plan 4: Management of air side economizer DSS interface Update suggestion base on new parameters Parameters to be changed by the user Select calculation method Suggestions for doing free cooling Time schedule of the suggestions

Action Plans (16/29) Action Plan 5: Scheduling the PV Maintenance General purpose Optimizing renewable energy production by detecting on time possible faults of the PV system. How does it work? By detecting the need for maintenance of the PV system and providing an alert prompting for appropriate maintenance actions. The identification of abnormalities and possible problems is facilitated through appropriate statistical methods. Indicators to be optimized Energy consumption Renewable energy production CO2 emissions

Theoretical Inference rules Action Plans (17/29) Action Plan 5: Scheduling the PV Maintenance Structure of the Action Plan Weather conditions RES Production Historical data Predicted data Energy Model DSS Application Theoretical Inference rules Sant Cugat Savona

Action Plans (18/29) Action Plan 5: Scheduling the PV Maintenance DSS interface Total historical and predicted production Click for details Alarm status of the PV system

Hourly differences between historical and predicted production Action Plans (19/29) Action Plan 5: Scheduling the PV Maintenance DSS interface Hourly differences between historical and predicted production Hour alert

Action Plans (20/29) Action Plan 6: Scheduling the sale/consumption of the electricity produced through the PV system General purpose Optimizing energy consumption or energy cost by exploiting RES production and load shifting techniques. Maximization of the self-consumption of electricity produced on-site, and selling of the surplus to make a profit. How does it work? By optimizing the selling/self-consumption of the electricity produced by a PV system. Different scenarios of energy market are considered (green strategy, finance strategy, peak strategy). Indicators to be optimized Income from the sale of surplus of energy produced through PV system Energy consumption Renewable energy production CO2 emissions

Theoretical Inference rules Action Plans (21/29) Action Plan 6: Scheduling the sale/consumption of the electricity produced through the PV system Structure of the Action Plan Weather forecasting Energy prices RES production Predicted data Prediction model RES production DSS Application Energy Model Theoretical Inference rules Sant Cugat Savona

Select calculation strategy Energy consumption, production and price Action Plans (22/29) Action Plan 6: Scheduling the sale/consumption of the electricity produced through the PV system DSS interface Select calculation strategy Energy consumption, production and price

Action Plans (23/29) Action Plan 6: Scheduling the sale/consumption of the electricity produced through the PV system DSS interface Suggestions to buy energy, sell energy and shift loads Click to show/hide graph Energy consumption, production and price

Action Plans (24/29) Action Plan 7: Scheduling the operation of heating and electricity systems towards energy cost optimization General purpose Minimize total energy cost of a building (or block of buildings) by optimizing simultaneously the operating schedule of its heating and electricity systems. How does it work? The real use of the infrastructures related with energy consumption and generation (PV fields, CHP systems and electricity storage) is simulated to specify based on the season (winter/summer) the schedule of the heating/cooling systems and then suggestions are made regarding when the energy generated by the systems of the buildings should be used, stored or sold in order to minimize energy cost or even make a surplus. In case that load shifting is possible, additional suggestions are made regarding when energy intensive processes should be scheduled.  Indicators to be optimized Energy cost

De-centralized sensor based Theoretical Inference rules Action Plans (25/29) Action Plan 7: Scheduling the operation of heating and electricity systems towards energy cost optimization Structure of the Action Plan De-centralized sensor based Weather forecasting Energy prices RES production Prediction model RES production Energy Model Electricity demand DSS Application Theoretical Inference rules Thermal demand Savona Energy prices

Action Plans (26/29) Action Plan 7: Scheduling the operation of heating and electricity systems towards energy cost optimization DSS interface Suggestions for load shifting Suggestions for battery use Suggestions for the operation of the thermal systems Energy flows. Choose day to display

Suggestions for load shifting Action Plans (27/29) Action Plan 7: Scheduling the operation of heating and electricity systems towards energy cost optimization DSS interface Suggestions for load shifting Remove load Add load

Suggestions for battery use Discharge the batteries Action Plans (28/29) Action Plan 7: Scheduling the operation of heating and electricity systems towards energy cost optimization DSS interface Suggestions for battery use Discharge the batteries Charge the batteries

Action Plans (29/29) Action Plan 7: Scheduling the operation of heating and electricity systems towards energy cost optimization DSS interface Suggestions for the operation of the thermal systems Energy to be produced by ThA Energy to be produced by ThB

Thank you for your attention! This project is co-funded by the European Union 57