DR in cold climate: Current R&D activities at IREQ Marie-Andrée Leduc, ing. M.Sc. Researcher Building Energetics IREQ
11/10/2014
1 www.peakload.org
Overview • Hydro-Québec and IREQ • Context • DR research • Residential • Space heating • Water heating
• Commercial and Institutionnal • HVAC control – Field study • Analytics
• Future work
11/10/2014
2 www.peakload.org
HYDRO-QUÉBEC AND IREQ
11/10/2014
3 www.peakload.org
Hydro-Québec at a glance • Integrated, gvnm’t owned utility • Installed capacity: 36 068 MW (hydro) • 33 885 km of transmission lines • interconnections with Ontario, NB/NE and NY markets • Annual electricity sales : 205 TWh • Customers: +4M • 3,7 M residential, • 291 000 commercial, 18 600 industrial
• AMI rollout to be completed by 2018 http://meters.hydroquebec.com/ 11/10/2014
4 www.peakload.org
IREQ – Institut de Recherche d’Hydro-Québec • 2 sites (IREQ, LTE) • 470 employees • Several fields of expertise including: End uses (Building Energetics Team) Electrical equipments Material science Robotics and civil works Mechanics, metallurgy and hydro/wind Information and measuring systems Electrical networks and mathematics 11/10/2014
5 www.peakload.org
CONTEXT
11/10/2014
6 www.peakload.org
Our climate Land of extremes
Montreal data 2013… • Ambiant temperatures
• Winter mean daily: -8,1°C (low: -24,7°C) • Summer mean daily: 23,2°C (high: 33,2°C)
• Water supply temperature • Winter 4°C • Summer 18°C
11/10/2014
7 www.peakload.org
Our peak • interruptible ind. (850 MW) • dual energy res. (640 MW)
• Electricity used as primary energy source for heating for most of our customers
11/10/2014
Source: Régie de l’énergie du Québec, R-3891-2014 and R-3864-2014
• 2014 Winter peak: 39 031 MW • Current DR programs :
8 www.peakload.org
DR RESEARCH
11/10/2014
9 www.peakload.org
DR applications under study • Focus on capacity
• Ancillary services are now being studied Existing programs or previous commercial pilots
IREQ’S role is to… Temperature sensitive Dual energy tariff • Estimate the potentiel of various DR resources • Evaluate different technical DR approaches (impacts, pros/cons) • Advice BUs using knowledge based on our research 11/10/2014
Source: NERC. Demand Response Availability Data System (DADS): Phase I and II Final Report. Princeton, NJ: North American Electric Reliability Corporation , 2011.
IREQ research
10 www.peakload.org
RESIDENTIAL customers Loads studied • • • • •
space heating water heating hot tubs ventilation appliances
Methodology • • • •
Characterization (extensive monitoring, tech. review) Simulation (ind. loads, load aggregation) DR potential evaluation Demonstration www.peakload.org
Residential Space heating Building stock characteristics • • • •
Wood-framed construction Heated basement and garage Electricity as primary source for heating Largely baseboard Zone by zone heating, decentralized control
Questions to be answered.. What is an acceptable setback for DR in a heating context? How do we define/measure acceptable ? Should preheating be done ? By how much ? How many thermostats should be controlled ? What will be the interactions between the different zones if partial control is done? How will it affects the DR potential what about the level of uncertainty of this resource ? What type of setpoint trajectories offers the best tradeoff between comfort and load reduction ? What about the rebound effect, how can it be limited ? What is the maximum allowable penetration rate of this DR resource ? ... 11/10/2014
12 www.peakload.org
Residential Space heating – lab testing EHBE: Experimental Houses for Building Energetics • unoccupied • better understand the physics (setpoint modulation, partial control) Fournier, M. Leduc, M-A, Study of Electrical Heating Setpoint Modulation Strategies for Residential Demand Response, eSim2014 Proceedings Le Bel, C., Gélinas, S., 2013. All-electric experimental twin houses: the ultimate demand management testing tool, Proceedings of The IASTED International Symposium on Power and Energy (PE 2013), Marina Del Rey, California, November 11-13, 2013. 11/10/2014
13 www.peakload.org
Residential Space heating – field study • occupied • bungalow 12 baseboards total and partial control several strategies tested
• load reductions whole house: 1,8 to 5,1 kW 4 thermostats: 0,9 to 2,6 kW
11/10/2014
14 www.peakload.org
Residential Space heating – building simulation and grid interation Grid impact of DR resource - simulated housing stock • Maximum allowable penetration rate depends on the shape of the grid’s overall demand profile and load control strategy
11/10/2014
Setpoint trajectories for optimized control Demand profiles for advanced strategies
15 www.peakload.org
Domestic Hot Water Heaters • DHWH stock characteristics Mostly customer owned Electric DHWH 93% Default setpoint 60°C Very cold water supply
• Research activities Load caracterization (lab testing and water withdraw profiles) Detailed load modeling (TRNSys) Grid integration modelling for DR potential evaluation Moreau, A., Control Strategy for Domestic Water Heaters during Peak Periods and its Impact on the Demand for Electricity, ICSGCE 2011, Energy Procedia 12 (2011) 1074-1082 11/10/2014
16 www.peakload.org
GIWH Innovation • Control of the lower element Standard tanks/thermostat retrofit capabilities
• DR potential 600 W/customer 5% less potential than shutting off power
• Benefits Maximizes hot water availability
• Proof of concept in SGZ – Boucherville (2 GIWH) 11/10/2014
17 www.peakload.org
Commercial and institutional (CI) customers Loads studied • HVAC
Research activities • DR potential evaluation Leduc, M.-A. (2013), Demand response in Quebec’s CI buildings: potential and strategies, ICEBO2013
• Simulation of HVAC controls optimization Lavigne, K., Daoud, A., Sansregret, S., Leduc, M.-A. (2014), Demand Response Strategies in a Small All-Electric Commercial Building in Quebec, Proceedings eSim2014
• Field study To be published at the Annual ASHRAE Conference 2015, abstract accepted
• Analytics • Baseline modeling www.peakload.org
CI HVAC control – Field study • Building characteristics
BANK
All-electric buildings DDC control
OFFICE
• DR strategies
RETAIL
o Customized DR strategies for each site o HVAC only, based on technical surveys, building simulation and analytics o Implemented into BAS, semi-ADR
Preconditionning Air intake limitation Zonal temperature adjustments Cycling *No ETS 11/10/2014
SCHOOL
RETAIL
19 www.peakload.org
CI HVAC control – Field study 50
100
40
80
30
60
kW
kW
o Jan-Feb 2014 o 4 to 6 events o Daily mean T -10°C to -24°C
120
60
20
40
10
20
0
0 0:00
2:00
4:00
6:00
8:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 0:00
profile
0:00
2:00
4:00
6:00
8:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 0:00
baseline
profile
baseline
• Significant load reduction AM
(31-66 %)
PM
• Acceptable impacts on comfort levels evaluation based on • Occupants surveys • BAS log data and additionnal sensors 11/10/2014
-6
-4
-2
0
2
W/pi2 20 www.peakload.org
CI – Analytics
2011 -9.23 (kW/°C) : (R2=0.94)
Classe 1, 31 jours
2012 -7.95 (kW/°C) : (R2=0.94)
Classe 1, 31 jours
40
500
500
450
450 Classe 2, 63 jours
40
Classe 2, 63 jours
400
450
450
400
400
400
350
350
350
10 20
350 20
30
10 20 300
Using 15-min data…
10/02/2013 24/02/2013
0 10 0
24/02/2013 10/03/2013
0 0
5 5
10/03/2013 24/03/2013
20
10 15 Classe 3, 33 jours Heure locale
20
5
250
0 0
5
10 15 Heure locale 250
20
10 15 Heure locale
20
Classe 4, 32 jours
200
Classe 3, 33 jours
40 24/03/2013 07/04/2013
21/04/2013 05/05/2013
kW
20 30
kW
kW
07/04/2013 21/04/2013
200
10 20
30 150 40 -40
-20
0 T (°C)
20
150 -40
40
-20
5
10 15 Heure locale
20
0 10 0
5
5
10 15 Classe 5, locale 44 jours Heure
20
0 0
10 15 Heure locale
20
0 0
5
10 15 Classe 6, locale 17 jours Heure
20
16/06/2013 30/06/2013
28/07/2013
40
150 -40
200
-20
0 T (°C)
20
40
150 -40
-20
0 T (°C)
20
40
kW
30 40
20 30
kW
kW kW
14/07/2013 28/07/2013
20
Classe 6, 17 jours
40
30 40
30/06/2013 14/07/2013
0 T (°C)
10 20
0 10 0
Classe 5, 44 jours
250
20 30
19/05/2013 02/06/2013
40
300
250
200
05/05/2013 19/05/2013
02/06/2013 16/06/2013
300
Classe 4, 32 jours
40
30 40
kW
le débutant Semaine le débutant Semaine
10 15 Heure locale
300
0 10 0
P (kW)
20 30
P (kW)
kW
30 40
kW P (kW) kW
kW
30 40
P (kW)
30/12/2012 13/01/2013
27/01/2013 10/02/2013
using non supervised clustering algorithms
2014 -8.07 (kW/°C) : (R2=0.85)
500
30/12/2012
13/01/2013 27/01/2013
• Operationnal patterns recognition
2013 -8.77 (kW/°C) : (R2=0.87)
500
10 20
20 30 10 20
D
L
M M J V Jour de la semaine
S
0 10 0
5
20
0 10 0
5
20
L
M M J V Jour de la semaine
S
5
10 15 Heure locale
20
0 0
10 15 Heure locale
D
0 0
10 15 Heure locale
5
10 15 Heure locale
20
Dominant response
• Linear regression models
FREQUENCY Null response
Negative response
Positive response
1
29/12/2013
12/01/14 0.8
16/02/14 12/01/2014
0.6
16/03/14 13/04/14
26/01/2014
0.4
11/05/14 0.2 15/06/14 1
23/02/2014
15
20
24
0
1
5
10
15
20
24
20/04/2014
50
1
-4
40
0.8
-6 -8 -10
30 20
-12
10
-14
0
1
5
10
15
20
24
K pos (kW/°C)
06/04/2014
-2
P base (x10kW)
K neg (kW/°C)
1
5
10
15
20
24
1
5
10
15
20
24
1
5
10
15
20
24
0.6 0.4 0.2
1
5
10
15
20
0
24
18/05/2014 1
0.8
10
0.8
0.6
5
0.6
2
R neg
15/06/2014
15
0.4
2
01/06/2014
1
R pos
Week starting
23/03/2014
04/05/2014
• VizBem – load profile vizualisation tool
10
MODEL PARAMETERS
09/03/2014
• Statistics (peak)
5
(°C)
(detailed thermal response)
09/02/2014
0
0.4
29/06/2014 0.2
-5
0.2
13/07/2014 Sun Mon Tue Wed Thu Fri Sat Pos
Nul
Nég
0
1
5
10
15 Time
20
24
-10
1
5
10
15 Time
20
24
0
Time
Millette, J.; Sansregret, S.; SIMEB: Simplified interface to DOE2 and EnergyPlus – A user’s perspective – case study of an existing building, BS2011 : 12e Conférence bisannuelle de l’IBPSA, Sydney, 11-14 novembre 2011. Sansregret, S., Lavigne, K., Spreadsheet tool development for visualizing building performance and simulation data to help calibrating models, 2014 ASHRAE/IBPSA-USA Building Simulation Conference, Atlanta, GA, September 10-12, 2014
Simeb, https://www.simeb.ca/ 11/10/2014
21 www.peakload.org
FUTURE WORK
11/10/2014
22 www.peakload.org
Ongoing DR research projects… • HEMS with focus on heating optimization Experimental validation of optimal partial control and complex setpoint trajectories
• Optimization of DR resources dispatch Optimally dispatching resources to achieve desired results, building on strengths and limits of each one
• Dynamic DR in the CI market Modeling predictive control Intelligent load manager based on extended submetering
• HQ buildings to be DR-ready • Autonomous load contribution to spinning and non-spinning reserves Smart loads
• Industrial DR activities DR potential evaluation Small industrial solutions 11/10/2014
23 www.peakload.org
Acknowledgement Eric Le Courtois, researcher LTE, project manager, Residential DR Ahmed Daoud, researcher LTE, project manager, CI DR Hugues Fortin, reseracher LTE, procect manager, Industrial DR • Stéphane Alarie, researcher IREQ • Jean-François Boudreau, researcher IREQ • Stéphane Boyer, technician LTE • Sylvain Chénard, technician LTE • Guy Desbiens, IT, IREQ • Michaël Fournier, researcher LTE • Nedeltcho Kandev, researcher LTE • Sylvain Lahaie, researcher LTE 11/10/2014
• • • • •
André Laperrière, researcher LTE François Laurencelle, researcher LTE Karine Lavigne, researcher LTE Marie-Andrée Leduc, researcher LTE Sylvie Legendre, Valorisation de la technologie • Sylvain Martel, researcher LTE • Alain Poulin, researcher LTE
24 www.peakload.org