the deployment of wireless networks in high voltage ... · 8 - basile l. agba epec 2012, london,...
TRANSCRIPT
11
The deployment of Wireless Networks in The deployment of Wireless Networks in High Voltage Substations: A feasibility High Voltage Substations: A feasibility
StudyStudy
Basile L. AGBA *, S. Riendeau, H. Bertrand, J. BélandBasile L. AGBA *, S. Riendeau, H. Bertrand, J. Béland
Scientist at IREQ and Associate Prof. at ETS (University of Québec, Canada)
Electrical Power & Energy Conference, EPEC 2012 – London, Canada, 10 – 12 october 2012
2 - Basile L. Agba EPEC 2012, London, Canada
AgendaAgenda
�� Introduction & study contextIntroduction & study context
�� Channel model and simulation parametersChannel model and simulation parameters−− Substation modelingSubstation modeling
−− Target technologies and simulation parametersTarget technologies and simulation parameters
�� Simulations results and DiscussionsSimulations results and Discussions−− Coverage resultsCoverage results
−− Throughput resultsThroughput results
−− Results discussions and recommendationsResults discussions and recommendations
�� Impact of impulsive noiseImpact of impulsive noise
�� Conclusion & perspectivesConclusion & perspectives
3 - Basile L. Agba EPEC 2012, London, Canada
Introduction & study contextIntroduction & study context
�� HydroHydro--Québec is the most important hydroQuébec is the most important hydro--electricity electricity provider in North Americaprovider in North America
−− 39 39 hydrohydro--electric power stationselectric power stations
−− 33 000 33 000 km transportation lines (km transportation lines (10 000 10 000 km @ km @ 735 735 kV)kV)
−− 505 505 substations and substations and 18 18 interconnections with USA and others interconnections with USA and others
provinces of Canadaprovinces of Canada
−− 110 000 110 000 km of distribution lineskm of distribution lines
�� HQ works actively to improve the interaction within the HQ works actively to improve the interaction within the grid with Smart Grid applicationsgrid with Smart Grid applications
�� The research institute, IREQThe research institute, IREQ
is dedicated for R&D activitiesis dedicated for R&D activities
Brief presentation of Hydro-Québec
4 - Basile L. Agba EPEC 2012, London, Canada
Introduction & study contextIntroduction & study context
What is particular for wireless in a substation ?
�� The wireless systems are affected by metallic The wireless systems are affected by metallic structuresstructures−− Multiple reflexionsMultiple reflexions
−− Diffractions and scatteringDiffractions and scattering
�� In addition, impulsive noise need to be take into In addition, impulsive noise need to be take into accountaccount
−− Partial discharges, corona effects and breaker operations are Partial discharges, corona effects and breaker operations are
some potential noise sources.some potential noise sources.
We proposed to study the feasibility of We proposed to study the feasibility of deploying WiFi, WiMAX and deploying WiFi, WiMAX and 900 900 MHz MHz systems in these specific conditionssystems in these specific conditions
5 - Basile L. Agba EPEC 2012, London, Canada
AgendaAgenda
�� Introduction & study contextIntroduction & study context
�� Channel model and simulation parametersChannel model and simulation parameters−− Global architecture and channel modelingGlobal architecture and channel modeling
−− Target technologies and simulation parametersTarget technologies and simulation parameters
�� Simulations results and DiscussionsSimulations results and Discussions−− Coverage resultsCoverage results
−− Throughput resultsThroughput results
−− Results discussions and recommendationsResults discussions and recommendations
�� Impact of impulsive noiseImpact of impulsive noise
�� Conclusion & perspectivesConclusion & perspectives
6 - Basile L. Agba EPEC 2012, London, Canada
Channel model & Simulation parametersChannel model & Simulation parameters
Global architecture
HV device
IntranetIP phone
High voltage(HV) device
7 - Basile L. Agba EPEC 2012, London, Canada
Channel model & Simulation parametersChannel model & Simulation parameters
Substation modeling
Concrete
walls
Lamps
Microwave
tower
TransformersBuildings
Concrete
walls
Lamps
Microwave
tower
TransformersBuildings
Concrete
walls
Lamps
Microwave
tower
TransformersBuildings
To reduce the computation time, only main structures are included
8 - Basile L. Agba EPEC 2012, London, Canada
Channel model & Simulation parametersChannel model & Simulation parameters
WiFi (IEEE WiFi (IEEE 802802..1111b/g/a) simulationb/g/a) simulation�� BS parameters:BS parameters:
−− EIRP allowed in each frequency band is used:EIRP allowed in each frequency band is used:•• 36 36 dBm in dBm in 22..4 4 GHz for GHz for 802802..11 11 b/g (Belair ARMb/g (Belair ARM33))•• 30 30 dBm in dBm in 5 5 GHz for GHz for 802802..1111a (Belair AP radio ERM a (Belair AP radio ERM 22))
�� CPE parameters:CPE parameters:−− A fixed CPE ( A fixed CPE ( 20 20 dBm of transmit power, dBm of transmit power, 6 6 dBi of antenna gain and dBi of antenna gain and 1 1 dB loss)dB loss)−− A mobile CPE (Cisco A mobile CPE (Cisco Unified IP Phone Unified IP Phone 7921 7921 G, G, with with 17 17 dBm of transmit power)dBm of transmit power)
WiMAX (IEEE WiMAX (IEEE 802802..1616e) simulatione) simulation�� BS parameters:BS parameters:
−− Airspan Airspan -- MicroMAX (MicroMAX (36 36 dBm of EIRP at dBm of EIRP at 55..8 8 GHz and GHz and 10 10 MHz bandwidth)MHz bandwidth)
�� CPE parameters:CPE parameters:−− AirspanAirspan--ProSTProST--2 2 ((20 20 dBm of transmit power, dBm of transmit power, 9 9 dBi antenna gain and dBi antenna gain and 1 1 db loss)db loss)
900 900 MHz Mesh simulationMHz Mesh simulation�� BS and CPE parameters:BS and CPE parameters:
−− 36 36 dBm of maximum EIRP:dBm of maximum EIRP:−− 900 900 MHz MHz –– FHSS: GEFHSS: GE--MDS radio, iNETMDS radio, iNET--II II 900900−− 900 900 MHz MHz –– OFDM: GEOFDM: GE--MDS radio, MercuryMDS radio, Mercury
9 - Basile L. Agba EPEC 2012, London, Canada
AgendaAgenda
�� Introduction & study contextIntroduction & study context
�� Channel model and simulation parametersChannel model and simulation parameters−− Global architecture and channel modelingGlobal architecture and channel modeling
−− Target technologies and simulation parametersTarget technologies and simulation parameters
�� Simulations results and DiscussionsSimulations results and Discussions−− Coverage resultsCoverage results
−− Throughput resultsThroughput results
−− Results discussions and recommendationsResults discussions and recommendations
�� Impact of impulsive noiseImpact of impulsive noise
�� Conclusion & perspectivesConclusion & perspectives
10 - Basile L. Agba EPEC 2012, London, Canada
Simulation results: CoverageSimulation results: Coverage
WiFi – 802.11 g (mobile CPE at 1.5 m AGL)
Client
802.11b
1 Mbps
802.11g
6 Mbps
802,11a
6 Mbps
Fixe (m) 1556 998 484Mobile (m) 764 490 238
Max. range corresponding to minimum data rate
11 - Basile L. Agba EPEC 2012, London, Canada
Simulation results: CoverageSimulation results: Coverage
WiMAX – 802.16e (CPE with 9 dBi gain, 20 dBm power)
12 - Basile L. Agba EPEC 2012, London, Canada
Simulation results: CoverageSimulation results: Coverage
900 MHz (GE-MDS CPE with OFDM)
13 - Basile L. Agba EPEC 2012, London, Canada
Simulation results: ThroughputSimulation results: Throughput
WiFi – 802.11 g (mobile CPE at 1.5 m AGL)
• 9 AP are used with independent channels(1, 6 and 11)
• The substation area is completely covered with 75 % above 18 Mbps and 25 % between 6 Mbps and 18 Mbps
14 - Basile L. Agba EPEC 2012, London, Canada
Simulation results: ThroughputSimulation results: Throughput
WiMAX – 802.16e (CPE with 9 dBi gain, 20 dBm power)
• 2 BS are used for optimization
• 40 % of the coverage area is above 12.7 Mbps and the minimum achievable rate is 3.2 Mbps.
15 - Basile L. Agba EPEC 2012, London, Canada
Simulation results: ThroughputSimulation results: Throughput
900 MHz (GE-MDS CPE with OFDM)
• 1 BS is used for 900 MHz system under OFDM modulation
• About 99 % of the coverage area is achieved with 7.2 Mbps using one BS
16 - Basile L. Agba EPEC 2012, London, Canada
Simulation results: DiscussionsSimulation results: Discussions
GOOD INSUFFICIENT
Global evaluation SATISFACTORY SATISFACTORY UNSATISFACTORY
GOOD N / A
Cybernetic security GOOD GOOD PASSABLE
EXCELLENT INSUFFICIENT
Data rate EXCELLENT EXCELLENT PASSABLE
WiMax 900 MHz
Propagation GOOD GOOD EXCELLENT
WiFi
Clients' number GOOD
Paquets' number GOOD
Other requirements GOOD
We make analysis based on the six criteria and we propose a synthesis with a qualitative assessment of each criterion:
• “Excellent” when the test is very good• “Good” when it is satisfactory• “Passable” when it is acceptable• “Insufficient” when it is unsatisfactory.
17 - Basile L. Agba EPEC 2012, London, Canada
AgendaAgenda
�� Introduction & study contextIntroduction & study context
�� Channel model and simulation parametersChannel model and simulation parameters−− Global architecture and channel modelingGlobal architecture and channel modeling
−− Target technologies and simulation parametersTarget technologies and simulation parameters
�� Simulations results and DiscussionsSimulations results and Discussions−− Coverage resultsCoverage results
−− Throughput resultsThroughput results
−− Results discussions and recommendationsResults discussions and recommendations
�� Impact of impulsive noiseImpact of impulsive noise
�� Conclusion & perspectivesConclusion & perspectives
18 - Basile L. Agba EPEC 2012, London, Canada
0.0104 0.0104 0.0104 0.0104 0.0105 0.0105 0.0105 0.0105 0.0105 0.0105-0.025
-0.02
-0.015
-0.01
-0.005
0
0.005
0.01
0.015
0.02
0.025
Time (s)
Am
plitu
de
Measured Noise over 100 µs
TIATA
max
Td
Sampling frequency: Sampling frequency: 1 1 GHzGHz
Frequency band: Frequency band: 200200--400 400 MHzMHz
� Amax: Max amplitude
� TIAT: Interval between pulses
� Td: Pulse duration
Impact of impulsive noise (onImpact of impulsive noise (on--going work)going work)
Many measurements are carried out in a substation inorder to derive an analytical model of HV environment
19 - Basile L. Agba EPEC 2012, London, Canada
10-1
10-2
10-3
10-4
10-5
0 204 8 12 16Eb/N0 (dB)
BER
Normal AWGN
receiver
“Impulsive”
receiver
Low impulsivecomponent
High impulsivecomponent
Impact of impulsive noise (onImpact of impulsive noise (on--going work)going work)
Knowing the channel characteristics in HV environment,we can design an optimised receiver
20 - Basile L. Agba EPEC 2012, London, Canada
AgendaAgenda
�� Introduction & study contextIntroduction & study context
�� Channel model and simulation parametersChannel model and simulation parameters−− Global architecture and channel modelingGlobal architecture and channel modeling
−− Target technologies and simulation parametersTarget technologies and simulation parameters
�� Simulations results and DiscussionsSimulations results and Discussions−− Coverage resultsCoverage results
−− Throughput resultsThroughput results
−− Results discussions and recommendationsResults discussions and recommendations
�� Impact of impulsive noiseImpact of impulsive noise
�� Conclusion & perspectivesConclusion & perspectives
21 - Basile L. Agba EPEC 2012, London, Canada
ConclusionsConclusions
�� The The 900 900 MHz technology stands out when considering MHz technology stands out when considering only the propagation aspect but the analysis has only the propagation aspect but the analysis has highlighted its inadequacy with some requirements such highlighted its inadequacy with some requirements such as the simultaneous number of customers, the as the simultaneous number of customers, the bandwidth limitation and the lack of interoperabilitybandwidth limitation and the lack of interoperability
�� The WiFi and WiMAX technologies are sufficient for The WiFi and WiMAX technologies are sufficient for these requirementsthese requirements
�� WiMAX seems to be the most costWiMAX seems to be the most cost--effective wireless effective wireless solution that meets the bandwidth need of HQ solution that meets the bandwidth need of HQ applications (In practice, a combination of technologies applications (In practice, a combination of technologies will be more relevant)will be more relevant)
�� The economic analysis showed that the number of AP or The economic analysis showed that the number of AP or BS required in a substation is the determining factor of BS required in a substation is the determining factor of the overall deployment costthe overall deployment cost
22 - Basile L. Agba EPEC 2012, London, Canada
The deployment of Wireless Networks in The deployment of Wireless Networks in High Voltage Substations: A feasibility High Voltage Substations: A feasibility
StudyStudy
Basile L. AGBA *, S. Riendeau, H. Bertrand, J. BélandBasile L. AGBA *, S. Riendeau, H. Bertrand, J. Béland
Contact : [email protected] or [email protected]