GreEn-ER Living Lab

A Green Building with Energy Aware Occupants

Benoit Delinchant

, Frédéric Wurtz

, Stéphane Ploix

, Jean-Luc Schanen

and Yves Marechal

G2ELab - Grenoble Electrical Engineering Laboratory, Grenoble University Alps, Grenoble, France

G-Scop Grenoble Industrial Engineering and Laboratory, Grenoble University Alps, Grenoble, France

Keywords: Green Building, Smart Building, Smart Micro-grid, Energy Efficiency, User Centred Building, Living Lab.

Abstract: In the context of doubling energy use worldwide by 2030, with 80% remaining carbon energy, and in which

buildings account for more than 40% of the total energy consumption, it is appropriate that buildings

contribute "intelligently" to reduce consumption, contribute to the increase of renewable energy production

and become a key node of the energy grid related to energy and transport of the eco-city. Moreover, these

“smart buildings” will be in the near future aggregating a huge amount of data through numerous sensors.

They will be connected to the neighbourhood, to the city and to the territory. Big data analytics and cloud

computing will bring new services to inhabitant and citizen. We are presenting in this paper, the smart building

“GreEn-ER”, which is a new building in the centre of the eco-city in Grenoble, France. It has been designed

with many “green” ideas and is a “living lab” supporting research and teaching in the field of sustainability.

This paper is focusing on the energy field and deals with electrical micro-grid interaction.

1 INTRODUCTION

1.1 The Building System

First of all, a building is composed by:

 a skin, which protects inhabitants from bad

external conditions (climatic, pollution, noise,

thieves…)

 systems to bring comfort (heating, ventilation,

air conditioning, lighting, cooking, sleeping,

entertainments, health care, security, …)

But we are not studing this building, we are

working on a complex system, the BUILDING

SYSTEM, which is composed by the previous

building in addition with:

 the building users (inhabitants, owner, energy

managers…)

 the building environment (weather, energy

grids, information network, transportation, …)

This system is really complex since it is subject to

many aspects such as:

 human (social, physiologic, phsycologic, etc)

 multi-physic (thermal, electric, mechanic, etc)

 economic (capex, opex)

 environmental (life cycle management, non-

renewable resources …)

These items are ones we have to adress in our

researches in the domain of energy in buildings.

1.2 Smart Building

Smart building is generally associated with home

automation but it is far more complex. One of the

challenges for energy based smart building success,

is to work together with at least three sectors:

The energy sector, which needs to anticipate the

consumptions and load shedding capabilities as well

as decentralized energy resources (DER) potential in

order to better manage production/consumption

correlation.

The building automation industry which must

move from classical BMS (building management

system) that are most coming from the industry

sector, to the city connected buildings and citizens.

The ICT sector which should collect the data of

buildings in order to feed service platforms in the

cloud for citizens, energy operators, and territorial

decision makers.

Heating building correspond in France to the main

consumption part as it can be seen with the strong

correlation of consumption with temperature (Figure

1).

316

Delinchant, B., Wurtz, F., Ploix, S., Schanen, J-L. and Marechal, Y.

GreEn-ER Living Lab - A Green Building with Energy Aware Occupants.

In Proceedings of the 5th International Conference on Smart Cities and Green ICT Systems (SMARTGREENS 2016), pages 316-323

ISBN: 978-989-758-184-7

Figure 1: Correlation between temperature and electrical

consumption in France – Gradient of daily consumption in

France in GWh/day as a function of the average temperature

in France (in °C).

Figure 2: Evolution tendency of residential energy

consumption /m² in France.

In Figure 2, while consumption reductions have

been obtainded for heating thanks insulation

improvement, the specific electricity is increasing a

lot. This part is nowadays a big challenge for passive

building where summer comfort becomes critical.

In Figure 3, we have put main topics around smart

building in a single picture and it is clear that smart

building is a complex system which has to deals with

many challenges.

1.3 Smart Building Challenges

It is possible to define several objectives for smart

buildings, according to social, environmental and

economical aspects.

 Sobriety, efficiency and energy optimization

 User comfort and services through internet of

things

 DER (Decentralized Energy Resource),

renewable energy produced

 Main contributor node of the city micro grid

multi-fluid energy management

 Mobility: link between Building – Transport

 New services associated with the Big Data

Analytics / Cloud computing

In Figure 3, we are defining several research

topics related to these challenges:

 Load management: load scheduling / adjusting

/ shedding, uncontrolled load prediction

Figure 3: What a smart building has to manage.

GWh/day

~50 GWh/day increase

per °C decrease

GreEn-ER Living Lab - A Green Building with Energy Aware Occupants

317

 Local or distributed multi energy sources and

storage management: arbitration between

generators, scheduling production / storage.

 Uncertainties management: regarding users

(equipment usage, comfort requierment),

regarding weather (heating/cooling needs,

renwable production), regarding energy price

 Local and distributed intelligence: energy and

comfort monitoring for inhabitants and service

providers, system actuation and control, data

mining, cloud computing.

 Life cycle management: adaptative modelling,

robust optimization, plug & play predictive

control…

Some of these challenges have been already

adressed by researchers and there is already a lot of

work in the litterature. For instance, starting from the

smart grid context, smart meter have been widelly

considered enought to make the grid smart but

Sharma has shown metering vulnerability and has

defined what is smart metrology meter (Sharma,

2015). In order to improve the knowledge about end-

users, load curve dissagregation must be applied on

the non-intrusive load-monitoring techniques.

Rowlands published recently a review and

recommendations on the end-user monitoring in order

to increase the measurments to loads, production and

storage. Numerous data are requiered to model load

demand at the level of the day. Torriti has made a

review of data and methods of time use studies such

as Markov chain technics (Torriti, 2015), while Fumo

has widelly use linear regressions but is claiming that

the increase of sensors will lead to individual models

instead of statistical ones (Fumo, 2015). Home energy

management systems (HEMS), based on such

modelling, enabled demand response in electricity

market, Khan review demand response programs in

various scenarios as well as incorporates various

architectures and models (Khan, 2015). Multi-agent

strategies are especially well suited in this building-

grid interaction or negociation (Labeodan, 2015).

Building energy management is not enough, users

comfort is critical to ensure sustainability engagment

of people. Shaikh conducted a state-of-the-art

intelligent control systems for energy and comfort

management in smart energy buildings (Shaikh,

2014).

Now, smart grid is dealing with ubiquitous

computing of smart building in which the home

environment is monitored by ambient intelligence to

provide context-aware services and facilitate remote

home control (Alam, 2012). A general trend for new

building is the nearly net zero energy buildings (Task

40/Annex 52, 2011). In Europe, the directive on

energy performance of buildings establishes the goal

of ‘nearly net zero energy buildings’ for all the new

buildings from 2020.

In France, all new buildings should comply with

energy positive by 2020. We are involved in

COMEPOS Project (www.comepos.fr) aiming at

constructing twenty five positive energy buildings in

France by 2018 in order to prepare this new

regulation. When energy generation is available in the

building neighbourhood it may become smarter since

it is possible to use more degrees of freedom. Lu et al.

has recently made a review on design optimization

and optimal control of grid-connected and standalone

nearly/net zero energy buildings (Lu, 2015).

In our team, our main challenges are to find

methodologies and to develop software for energetic

systems design and operation in their environment,

and during life cycle. This includes:

 the optimal design with operating costs (Capex

+ Opex).

 the optimal operation of consumption/

production/storage.

 “human in the loop”, to define comfort/cost

trade-off, to give sobriety advices…

In the following parts of this paper, it will be

presented through our experimental platform how we

are dealing with these challenges through the

following activities:

 Modeling: systemic approach, multicriteria

tradeoff, scalability and uncertainty

 Optimization: dedicated algorithms and

strategies

 Smart Ubiquity: information network, local

and distributed computing

 Transdisciplinary Approaches : working with:

sociologist, economists, computer scientists

2 ENERGY EFFICIENCY

2.1 GreEn-ER Building

GreEn-ER is a new building in Grenoble, dedicated

to develop creativity, entrepreneurial spirit, and

sustainability popularization in an environment

combining training students, researchers, and end

users. GreEn-ER is hosting master level training, for

students of “Energy, Water and Environmental

Engineering School” (Grenoble INP ENSE3).

Some figures:

 A 6 floors building with 4500 m² space per

floor for platforms teaching / research

SMARTGREENS 2016 - 5th International Conference on Smart Cities and Green ICT Systems

318

 2000 people welcomed in the building,

including 1,500 students.

 1 research laboratory

 2 restaurants (a brasserie and a university

restaurant)

 500m

of space at the Library

Figure 4: GreEn-ER: a building for energy learning and

research.

2.2 Designed with Green Ideas

Many green ideas are in experiment, such as energy

recovery. Data server room calories are reused to heat

an atrium (Figure 5). Free cooling strategies using

natural convection through the atrium (Figure 6), and

forced convection in offices.

Figure 5: Server room calories recovery for atrium heating.

Figure 6: Natural ventilation through the atrium.

Dual flow ventilation with high efficiency

recovery and low temperature supply (Figure 7):

 heating: 30/25°C (occupation/inoccupation)

 cooling: 19/23°C (occupation/inoccupation)

Figure 7: HVAC system.

And some other sustainable solutions:

 Automation: HVAC, lighting and blinding are

based on local sensors of temperature,

luminosity, CO2, and occupancy.

 Energy Sobriety: limited air conditioning, no

hot water in toilets. Power supply switching

 Water Sobriety: 40% of the water consumption

is from rainwater.

 Green Roof: free spaces on the roof are covered

by vegetal plants.

Total primary energy consumption will be less

than 2200 MWh / year which correspond to 110

kWh/m

2.3 Energy Autonomy and Micro Grid

In GreEn-ER, PREDIS-MHI is 600 m² platform

energy systems (

Figure 8

), has been specifically

designed to reach zero energy building, and to study

building or neighbourhood autonomy.

Figure 8: PREDIS MHI, production and storage.

20kW of photovoltaic panels installed on vehicles

roof, and other are planned to be installed on the

building roof. Other electrical productions are

available in PREDIS platform such as a fuel cell and

combined heat & power (CHP) which is also able to

heat our platform. Storage capabilities have also been

GreEn-ER Living Lab - A Green Building with Energy Aware Occupants

319

installed with electrical vehicles, and laptop rooms. A

50kWh stationary battery will be added.

Figure 9: PREDIS MHI micro grid structure.

LC: Local Controller, CC : Central Controller, LM : Local

Monitor.

Optimal control solutions based on predictive

models will be tested in this platform. Solving the

problem of demand response requires determining a

generation and a controllable load demand policy that

minimizes, over a planning horizon, an objective

function subject to economic and technical

constraints. This policy is used as reference for the

voltage and frequency control in microgrid real-time

operation.

Load demand can be classified by priority and

type as critical:

 Critical Load Demand: has to be full supplied

all the time, otherwise, it will cause deficit in

the system.

 Reschedulable Load Demand: has a particular

characteristic of being able to be allocated

across a range of time.

 Curtailable (shedable) Load Demand: may

have the power supply cut, as a non-priority

load, if necessary.

 Diffuse Load Demand: is a new concept made

to deal with a thermic load demand, having the

diffuse effect or the pre-diffuse behavior.

We are solving such an energy management using

a deterministic mixed-integer linear programming

problem, where the planning horizon is 24 hours with

one-minute time steps (Tenfen 2014).

2.4 Real Time Energy Management

In our platform, several hundred measuring points

and control have been set up such as HVAC,

dimmable lighting, blind, electrical plug consumption

measure and switch…

These measures and commands are accessible

through the building network infrastructure and the

internet. Its communication protocol is web service

based enabling interoperability.

Many devices are added from the delivered

building such as wireless sensors (433MHz, ZigBee,

EnOcean, DeltaDore). In SmartGreen 2014, Abras

has presented the interoperability framework that we

have developed in order to manage this

interoperability through web-services (Figure 10,

Abras, 2014)

Figure 10: Sensor/actuators interoperability framework.

In 2013, Dang has presented the optimal control

of laptops charging/discharging to fit photovoltaic

production (Figure 11, Dang, 2013). This experiment

was done in our previous building and will be

deployed in the new one since it allows load shedding

and photovoltaic energy storage.

Figure 11: Real time control of 15 laptops

charging/discharging for demand response.

In this automation of energy management, users /

occupants are integrated in the loop. At this moment,

occupants are able to define set points such as

temperature, CO2 concentration, luminosity, heating

mode. Moreover, their feedback is required and

individual comfort will be learned progressively.

SMARTGREENS 2016 - 5th International Conference on Smart Cities and Green ICT Systems

320

Beside this automation, many things can be done

to help users to understand their own consumption

and to let him acting on the system in order to learn

about it. It will be discussed further in the part 3.

2.5 Multi-flow Energy Optimization

Both electric and thermal energy flow must be taken

into account in smart buildings since one of the main

consumption in building is HVAC (heating,

ventilation and air conditioning). Predictive models

for both electrical and thermal equipments are then

required for the optimal operation.

Regarding thermal behaviour of buildings,

dynamic simulation are done based on thermal

properties of the skin (Figure 12), depending on

weather conditions, occupancy and control set points.

The results are the energy consumption with peak

power, and occupant comfort criteria.

Figure 12: Thermal dynamic simulation of GreEn-ER

(COMFIE software).

The previous software (COMFIE) is mainly used

for design stage (not in control), since occupancy is

predefined by scenario. In order to make a control

depending on the occupant behaviour, we are able to

make a co-simulation with this software and agent

based modelling occupants (Gaaloul, 2013). This co-

simulation is allowed thanks to our interoperability

framework based on software component

(ICAr/MUSE, http://muse-component.org).

Moreover, predictive control is done using

optimization algorithms and requires dedicated

modelling. The electric equivalent circuit method is

recognized to be a good compromise between pure

physical modelling which are too expensive and

require too many information, and black box model

based on mathematics which are not enough robust

regarding prediction uncertainties. Robust self

learning technics of physical equivalent circuit have

been developed (Le Mounier, 2014).

There are Optimization technics using mixed

linear integer programming (MILP) which needs

linear modelling. These technics are especially

efficient for smart grid and microgrid (Tenfen, 2014)

but have limitation for non linear models which have

to be discretized (Le Mounier, 2014). The difficulty

in using non-linear model is to compute partial

derivatives in order to couple it with gradient based

optimization algorithms. We are using two different

technics, the automatic differentiation (Dinh, 2015),

and the adjoin method (Artiges, 2015), depending on

the nature of the model.

Energy optimization is also strongly related to

measurements. In (Artiges, 2015), we are able to

control energy system but also to determine which

sensors are best suited to perform this control.

3 GREEN-ER: A LIVING LAB

3.1 The Living Lab Concept

A definition is given by the European Network of

Living Labs (ENoLL: www.openlivinglabs.eu),

which is “a real-life test and experimentation

environment where users and producers co-create

innovations”.

Academic institutions such as Harvard, Yale and

Cambridge have adopted this concept within their

strategy for sustainability (Graczyk, 2015). In most

living lab concepts, the key components have been

identified to be the ICT & Infrastructure,

management, partners & users, research.

ENoLL defines four main activities:

 Co-creation: co-design by users and producers

 Exploration: discovering emerging usages,

behaviours and market opportunities

 Experimentation: implementing live scenarios

within communities of users

 Evaluation: assessment of concepts, products

and services according to socio-ergonomic,

socio-cognitive and socio-economic criteria.

3.2 AmiQual4Home Project

GreEn-ER and more especially the PREDIS platform

is involved AmiQual4Home project which stands for

Ambiant Intelligence for Quality of Life. It is an

Innovation Factory which is an open research facility

for innovation and experimentation with human-

centred services based on the use of large-scale

deployment of interconnected digital devices capable

of perception, action, interaction and communication.

GreEn-ER Living Lab - A Green Building with Energy Aware Occupants

321

The Amiqual4home Innovation Factory is a

unique combination of three different innovation

instruments:

 Workshops for rapid prototyping of devices

that embed perception, action, interaction and

communication in ordinary objects,

 Facilities for real-world test and evaluation of

devices and services,

 Resources for assisting students, researchers,

entrepreneurs and industrial partners in

creating new economic activities.

3.3 Workshops for Co-creation and

Innovation

We have defined several workshops for users in order

to develop ideas, innovative products & solutions.

Each workshop is a set of tools available in the lab.

 Brainstorming: to produce innovative ideas for

the developing of new services.

 Role Playing: to test how users are interacting

together or with the system (a person may

simulate the response of the system).

 Wizard of OZ Prototyping: to simulate new user-

specialized interfaces to monitor and control

things.

 Sensors: to develop and test new sensors

technology, positioning and usefulness.

Interaction of sensors with occupants.

 Integration of Industrial Solutions: to test existing

energy and control systems by integration in our

platform in order to study the interaction with

other systems in the whole building.

 Data Mining: to discover patterns in large data

sets in order to discover trends and behaviour of

smart buildings usages.

 Data Processing: to develop algorithms that use

data and transform it into actions or meaningful

information.

 Simulation: software for building simulation

(energy, human behaviour, control …).

 Adaptive Modelling: development of algorithm to

fit predictive models and data in order to prevent

uncertainties and improve robustness.

Some workshops can be already targeted such as:

 User Monitoring: Analysis of user behaviour and

comfort

 Smart Grid Monitoring: Analysis of electricity

exchanges between building and distribution grid.

 Demand Side Management: Development of

solution for participating in grid support strategies

such as demand response, load shedding, etc.

These workshops are now in development and

will be sources for future research publications. These

workshops are end-user side view and will merge

smart grid requirements (such as energy peak

reduction, decentralized renewable energy

integration), and end-user requirements (such as

comfort, economy and sustainability).

4 CONCLUSIONS

The energy context and the potential of buildings in

order to improve sustainability are introduced in this

paper. GreEn-ER is then presented and the energy

efficiency of the building is highlight for its design as

well as for its operation. This building is presented as

a support for research in the domain of energy and

sustainability. We have presented scientific results

based on previous publications that have been

realized through this platform, such as energy

autonomy of building or micro-grid, real time energy

management, multi-flow energy optimization.

But GreEn-ER, with the PREDIS MHI platform,

is now becoming a “Living lab”. Building users are

researchers and students and they are able to improve

their own environment through several workshops

which have been presented in this paper. Thanks to

these tools, users can do innovations by themselves.

This interaction between the building and users are

our experimental field for the near future.

ACKNOWLEDGEMENTS

This work was supported by Grenoble-INP ENSE

IRT Nanoelec, Equipex Amiqual4Home (ANR-11-

EQPX-0002), and research project PRECCISION

(ANR-12-VBDU-0006).

REFERENCES

Sharma, Konark; Saini, Lalit Mohan, “Performance

analysis of smart metering for smart grid: An

overview”, Renewable & Sustainable Energy Reviews,

vol.: 49 pp.720-735, SEP 2015.

Torriti, Jacopo, “A review of time use models of residential

electricity demand“, Renewable & Sustainable Energy

Reviews, Volume: 37 Pages: 265-272, SEP 2014.

Fumo, Nelson; Biswas, M. A. Rafe, “Regression analysis

for prediction of residential energy consumption”,

Renewable & Sustainable Energy Reviews, Volume:

47, Pages: 332-343 JUL 2015.

SMARTGREENS 2016 - 5th International Conference on Smart Cities and Green ICT Systems

322

Khan, Aftab Ahmed; Razzaq, Sohail; Khan, Asadullah; et

al. “HEMSs and enabled demand response in electricity

market: An overview”, Renewable & Sustainable

Energy Reviews, Volume: 42, Pages: 773-785, FEB

2015.

Labeodan, Timilehin; Aduda, Kennedy; Boxem, Gert; et al

“On the application of multi-agent systems in buildings

for improved building operations, performance and

smart grid interaction - A survey”, Renewable &

Sustainable Energy Reviews, vol 50, pp. 1405-1414,

OCT 2015.

Shaikh P.H., Nor N.B.M., Nallagownden P., Elamvazuthi

I., Ibrahim T., “A review on optimized control systems

for building energy and comfort management of smart

sustainable buildings”, Renew Sustain Energy Rev, vol

34, pp. 409–429, 2014.

Alam, Muhammad Raisul; Reaz, Mamun Bin Ibne; Ali,

Mohd Alauddin Mohd, “A Review of Smart Homes-

Past, Present, and Future”, Ieee Transactions On

Systems Man And Cybernetics Part C-Applications And

Reviews, vol. 42, Issue: 6, pp. 1190-1203, NOV 2012.

Task 40/Annex 52, Anna Joanna Marsza, Per Heiselberg,”

Zero Energy Building definition - a literature review”.

A technical report of subtask A, 2011.

Lu, Y.; Wang, S.; Shan, K., “Design optimization and

optimal control of grid-connected and standalone

nearly/net zero energy buildings”, APPLIED

ENERGY, vol.: 155, pp.463-477, OCT 2015.

Tenfen D.; Delinchant B.; Wurtz F.; Finardi E. C.; Rolim

J.; Fernandes Rubipiara C., “Load Demand, Batteries,

and Electric Vehicles Modelling to the Energy

Management of Microgrids”, Workshop ELECON,

Magdeburg, October 2014.

S. Abras T. Calmant S. Ploix D. Donsez F. Wurtz, O. Gattaz

B. Delinchant « Developing Dynamic Heterogeneous

Environments in Smart Building Using iPOPO »

SmartGreen’14, International Conference on Smart

Grids and green IT Systems, Barcelona, Spain, 3:4

April 2014.

Dang H.-A., Delinchant B., and Wurtz F., “Toward

autonomous photovoltaic building energy

management: modeling and control of electrochemical

batteries”, Proceedings of Building Simulation 2013,

IBPSA’13, Chambery, August 2013.

Gaaloul S., Dang H.-A., Kashif A., Delinchant B. and

Wurtz F., “A new co-simulation architecture for mixing

dynamic building simulation and agent oriented

approach for users behaviour modelling”, Proceedings

of Building Simulation 2013: 13th Conference of IBPSA

2013, Chambery, August 2013.

Le Mounier A., Delinchant B., Ploix S., “Determination of

relevant model structures for self-learning energy

management system”, BSO’14, Building Simulation

and Optimization Conference, 23-24 June, London,

England, 2014.

Artiges N, Nassiopoulos A., Vial F., Delinchant B.,

“Optimal Energy Management Using Model Predictive

Control: Application on an Experimental Building”,

Climamed’15.

Dinh V.-B., Delinchant B., and Wurtz F., “Optimal sizing

of a complex energy system integrating management

strategies for a grid-connected building”, Building

Simulation Conference, Hyderabad in December 7-9,

2015.

Graczyk G., “Embedding a Living Lab approach at the

University of Edinburgh,” The University of

Edinburgh, Jun. 2015.

GreEn-ER Living Lab - A Green Building with Energy Aware Occupants

323