Reference Network Models

Reference Network Models
Dr. Carlos Mateo & Dr. José Pablo Chaves-Ávila
Institute for Research in Technology – IIT. Comillas University (Madrid)
ISGAN Academy
March 3rd, 2017

Table of contents
1. Introduction to (the need) of reference
network models
2. (How to) build reference networks
3. Case studies (and examples)
4. Further work
5. Q&A

Why Reference Network Models?
• The integration of new Smart Technologies in
power systems requires…
• Technical (engineering based) and economic analyses
• Short term / operation: power flows (electricity) and
communications
• Mid-long term / planning: decide optimal reinforcements and
new infrastructure
• How scalable are these results?
• How replicable are these results?
• Need of typical/representative distribution
networks
• Include medium and low voltage
• Based on regional standards and regulation

Definition and uses
• “A reference or representative network is a theoretical network that can
be used as an objective reference for actual grids”.
• Alternatives:
• reference network models (RNMs) or norm models: RNMs or norm
models are large-scale optimization models which rely on
geographical information
• provide regulators with an estimation of the efficient costs that would be incurred by a
distribution company supplying a certain geographical area
• Eg. Spain, Sweden and Chile.
• representative networks/feeders: a type of reference network that
may rely or not on geographical information and can be considered
appropriate to describe behaviour of a set or cluster of real
distribution feeders
• mainly proposed to benchmark continuity of supply
• or, regulate overall distribution costs as in Brazil
• study future network concepts: UK, Italy, EU-funded projects
• Standard networks: IEEE, CIGRE, …

Revenue cap formula in Spain (2008)
dc: Distribution company
n: Year
Rn
dc
: Allowed revenue
IAn: Adjustment taking into account price indexes
Yn-1
dc
: Allowed change in revenue to take into account new demand or
additional connections.
Qn-1
dc
, Pn-1
dc
: Incentive or penalty for complying with energy losses and
continuity of supply targets.
   1 2 2 1 1 11dc dc dc dc dc dc dc
n n n n n n n nR R Q P IA Y Q P            
There is a base revenue, an incentive for losses and quality of service, price indexes,
and an update of the revenue every year.
The amount of information to calculate the increase of revenues increases
significantly, reducing the asymmetry of information between the regulator and
the distribution companies.
Reference Network Models: a tool to
reduce information asymmetry

Reference Network Models: versions
• Greenfield
 It plans the whole distribution network to connect consumers
and distributed generation to the transmission substations,
without considering the existing network, but taking into
account the same technical constraints and the same planning
criteria.
 It can be useful to support in the calculation of the base
revenue (Rn
dc)
• Brownfield
 It expands the network to connect additional vertical and
horizontal demand, given the initial network, and considering
the same technical constraints and planning criteria.
 It can be useful to support in the calculation of the allowed
change in revenue to take into account new demand or
additional connections (Yn-1
dc)

Main features
Highlights
• The models are large-scale, being able to deal with millions of customers.
• The models plan the network installations required (with their technical parameters) and their cost.
• They design low, medium and high voltage networks.
• They model urban and rural areas.
• The objective of the models is to find the most cost-efficient solution that is able to supply the
demand and connect the distributed generation.
• They take into account technical constraints (current and voltage limits), geographical constraints
and reliability targets.
Main Inputs
 Power and GPS location of every single consumer, new or existant, as well as the corresponding
information for distributed generators.
 Catalogue of equipment and parameters.
Main outputs
 Cost efficiently incurred required to build a network.
 Topology of the output network

Large-scale Distribution Grid Planning
(RNM)
> 10
> 102
>
106
> 104
• Input Data: HV, MV and LV customers, and transmission
substations
•Results of the model: CAPEX (lines, transformers, substations)
and OPEX

Inputs
 Georeferred consumers and transmission
substations
 Coordinates and demand
 Catalogue of standard equipment
 Substations, electrical lines, capacitor banks, …
 Maintenance crews, equipment to improve reliability,…
 Technical and economic parameters
 Discount rate, demand growth, simultaneity factor, loss factors, …
 Geographical data
 Topography, nature reserves, lakes, …

Location of consumers
• When the location of consumers is unknown
there is an external module to identify
coordinates of consumers.
Input:
Street map image
Identifying
streets and parks
Placing consumers

Planning steps
• Planning is applied first to low voltage, then
to medium voltage and finally to high voltage.
• In high voltage the location of the
transmission substations is also an input to
the model.
Input:
Customers
Planning the
MV/LV transformers
Planning the
LV feeders

Geographical constraints and
planning models
GIS
Electrical lines path
Location cost
Transmission substation
HV/MV Substation
MV/LV Transformers
LV consumers
HV Network
MV Network
LV network
Planning models
A
B

Automatic settlement identification

Urban areas
• The models are able to plan networks in urban areas.
• Inside settlements feeders are constrained by a street
map which is automatically calculated by the model
(i.e. the feeder do not cross through buildings).

Street map impact
(1) With street map (2) Without street
map
Low voltage network 1418km 1214km
Medium voltage network 851km 619km
(1) With street map (2) Without street map

Rural areas
• The models are able to plan networks in rural areas,
with several farms and settlements.
• Outside settlements networks are constrained by
forbidden ways through (such as lakes and natural
reserves) and depend on topography.

Geographical constraints
Topography (level curves)
Nature reserves, lakes,…
Nature
reserve

Source: Jenkins (2014)
Reference Network Model (detail)

Example distribution networks
Thick colored lines represent MV feeders.
Thick black lines are loops in the MV network (normally open)
Thin black lines are LV feeders
MV/LV transformers are represented by red circles
Consumers are identified with small dots.
The blue triangle represents the HV/MV substation.
Five MV feeders in an urban area inside a
settlement.
Semi-urban configuration, modeling the outskirts of
a city.
Source: DSO Observatory

Adapting the RNM to US grids
Original
EPRI Test Feeder J
RNM-US Version
U.S. catalog
Phase A
Phase B
Phase C
Two-phase
Three-phase
Source: Project Smart-DS, Synthetic models for advanced,
realistic testing of distribution systems and scenarios.
Financed by: U.S. Department of Energy’s Advanced
Research Projects Agency-Energy (ARPA-E)

Case studies
• What is the impact of increasing share of DERs:
generation (PV, wind, CHP), storage, demand
response, electric vehicles, on distribution network
investment ?
• How different alternatives for location and operation
of DERs may affect the operation and planning of
distribution networks ?

Typical methodology
Build
Initial
Network
Customers&DG
Substations
DER
- Voltage level
- Location
- Hourly profile
Expand
the
initial
network
Initial
Network
Select
Scenario
LV feeders
Assess
distribution
costs
Expansion
Network
MV/LV
Transformers
MV feeders

Case studies
• SOLAR PV
• Increasing penetration of solar PV in distribution networks (due
to the lack of simultaneity with peak consumption) may increase
the needs for network reinforcements
• ENERGY EFFICIENCY AND DEMAND RESPONSE
• Active demand in the form of energy and peak consumption
reductions may decrease the level of utilization of distribution
networks and postpone the need for future reinforcements
• ELECTRIC VEHICLE (EV) CHARGING
• Smart EV charging during off-peak hours would avoid the need
for distribution network reinforcements even with high EV
penetration levels
• RURAL ELECTRIFICATION
• RNM as a supportive tool for rural electrification decisions

3% Penetration
Source: MIT Solar Study
Impact of solar PV in distribution
networks (RNM)

8% Penetration
networks (RNM)

11% Penetration
networks (RNM)

14 % Penetration
networks (RNM)

17% Penetration
networks (RNM)

17 % penetration
networks (RNM)
• Distribution network reinforcements & their associated costs can be
significant

Case studies
• SOLAR PV
• Increasing penetration of solar PV in distribution networks (due to the
lack of simultaneity with peak consumption) may increase the needs for
network reinforcements
• Active demand in the form of energy and peak consumption reductions
may decrease the level of utilization of distribution networks and postpone
the need for future reinforcements
• Smart EV charging during off-peak hours would avoid the need for
distribution network reinforcements even with high EV penetration levels

RNM Active Demand (AD) impact on
network costs
Developed by IIT for EU ADVANCED Project
Input OutputModel 1 Model 2
• Orography
• Peak Demand
• Electrical limits
• Std Equipment
Greenfield
ReferenceNetworkModel
Incremental
ReferenceNetworkModel
•Reinforcements
•Quality of supply
•Energy losses
• New peak load
• AD programs

AD= energy efficiency + demand
response
% ∆ Energy (Energy
Efficiency)
% ∆ Peak (Demand
Response)
0,4
0,5
0,6
0,7
0,8
0,9
1
1 3 5 7 9 11 13 15 17 19 21 23
No Active Demand (baseline)
Informative bill
In-Home Display (IHD)
Website
0,4
0,6
0,8
1
1 3 5 7 9 11 13 15 17 19 21 23
Time of Use (no-automation)
Time of Use (automated)
Critical peak (no-automation)
Critical peak (automated)
Real-time pricing (no-automation)
Real-time pricing (automated)
No Active Demand (baseline)

Impact of AD on distribution networks
No AD Maximum
Potential of
AD
Urban network
Source: ADVANCED Project

Impact of AD on distribution networks
No AD Maximum
Potential of AD
Rural network

AD avoiding distribution network
investments
5.3% 5.3%
6.6% 6.6%
4.1%
4.8%
5.1%
5.7%
0%
1%
2%
3%
4%
5%
6%
7%
DynamicPricing Feedback DynamicPricing Feedback
Rural Urban
%initialcostofthenetwork
NoAD Baseline Optimistic Tec. Pot. %:No AD %: Optimistic
Network Reinforcements needed

Case studies
• SOLAR PV
• Increasing penetration of solar PV in distribution networks (due to the
lack of simultaneity with peak consumption) may increase the needs for
network reinforcements
• Active demand in the form of energy and peak consumption reductions
may decrease the level of utilization of distribution networks and
postpone the need for future reinforcements
• Smart EV charging during off-peak hours would avoid the need for
distribution network reinforcements even with high EV penetration levels

EV penetration study: distribution
networks
Developed by IIT for EU MERGEProject
Number of
Customers
Power of
Customers (kW)
LV
Feeders
MV
Feeders
MV/LV
Transf.
HV/MV
Subs.
LV MV LV MV km km Number Number
Tourist 154,984 15,171 816,663 204,538 1,058 600 1,089 7
Rural 25,637 921 120,987 41,293 378 567 267 3
New City 106,978 197 564,913 3,133 678 780 838 13
Big City Old Network 8,173 212 53,785 1,638 31 60 93 3
Big City New Network 34,567 355 227,004 4,958 313 285 412 2
Greek City 38,737 66 179,838 5,450 246 227 318 1
• Tourist area
• Rural area

0 5 10 15 20
-0.5
0
0.5
1
1.5
2
2.5
3
3.5
x 10
5
Hour
Power[kW]
Aggregated
0 5 10 15 20
-0.5
0
0.5
1
1.5
2
2.5
3
3.5
4
x 10
5
Hour
Power[kW]
Aggregated
0 5 10 15 20
-0.5
0
0.5
1
1.5
2
2.5
3
3.5
x 10
5
Hour
Power[kW]
Aggregated
EV charging strategies
Dumb Charging
(peak charging)
ToU Tariffs
(low prices at off-peak)
Smart
Charging
Residencial load
Industrial load
EV charge
Source: EU MERGE Project

• In an scenario with high penetration of EVs (2030)
dumb charging would imply huge incremental
additions of network reinforcement
Distribution network investment:
EV dumb charging
0 % 10% 20% 30% 40% 50% 60%
Rural
Tourist
Big city old network
Big city new network
New city
Percent Reinforcements
LV Feeders
MV/LV Transformer Substations
MV Feeders

• Incremental investment in MV/LV
transformers and LV networks are drastically
reduced with EV charging at off-peak hours
Distribution network investment:
EV charging strategies
Smart Valley Peak
0 %
5 %
10%
15%
20%
25%
30%
35%
40%
PercentReinforcements
New city
Tourist
Rural
Smart Valley Peak
0 %
5 %
10%
15%
20%
25%
30%
35%
40%
PercentReinforcements
New city
Tourist
Rural

Sensitivities
The Brownfield RNM can be
applied to obtain sensitivities of
the reinforcement cost to input
parameters, such as distributed
generation or electric vehicle
penetration.

• The Reference Electrification Model
(REM) is a large-scale rural
electrification tool that provides
assistance in establishing an
electrification planning.
• REM uses heuristic algorithms in
order to minimize total cost. Both
financial costs and social costs are
included.
• This model uses RNM to calculate
the grid extension designs and the
networks of the microgrids.
• REM has been applied to vast areas
such as the Vaishali District in Bihar
(India).
Reference Electrification Model

Extension 11kV Extension 400V Stand-AloneMicrogrid 11kV Microgrid 400V
REM provides a graphical representation of the networks that RNM
designs in Google Earth files.
Reference electrification Model

Conclusions
• The networks built depend on sizing criteria, continuity of supply targets, the
catalogue of standard equipment, technical and economic parameters:
• These criteria are necessary to run the model and obtain a reference
network
• Results depend, therefore, on the different criteria up to a certain
extent.
• Differences between real networks and reference network depend on such
criteria, but results use to be similar.
• A reference network model is a tool intended to support the regulator.
• It helps to reduce the asymmetry of information between the regulator and
the distribution system operators.
• It is a tool which is also useful for network analysis, to estimate network cost,
distributed generation impact, energy losses and quality of service.
• The main objective is to estimate the costs efficiently incurred.
• Models are also useful to build synthetic networks.

• SYNEX, SYSTEP (Consultant companies): Building Chilean networks for studying the value-added
distribution
• ORMAZABAL (Electric manufacturer): Urban, semi-urban and rural networks to make a cost-
benefit assessment of storage
• CNE (Spanish regulator): Planning the distribution system in Spain based on real consumer
coordinates, covering 504,645 km2 and 27 million customers
• MERGE (EU project): Building 5 distribution networks in Spain based on real consumer
coordinates and 1 synthetic network in Greece, up to 155k consumers
• ADVANCED (EU project): Building distribution networks in Italy, Spain, France and Germany in
order to analyze demand response.
• IMPROGRESS (EU project): Case studies in The Netherlands, Germany and Spain to analyze the
impact of DG.
• FUTURE OF SOLAR (MIT project): 6 distribution network case studies to analyze the future role
of solar PV
• DSO OBSERVATORY (JRC – European Commission): Building 13 European representative
distribution networks, including MV and LV feeders downstream a primary substation.
• UTILITY OF THE FUTURE (MIT): Analyzing the impact of distributed energy resources and
distribution network charges.
• SMART-DS (U.S. Department of Energy): Developing a U.S. RNM and building synthetic networks.
Some relevant projects where Reference
Network Models have been used

Home work
• Build a full distribution network (MV/LV)
using a web based tool:
• Set initial parameters.
• Build the optimal distribution network.
• Analyze technical results
• Analyze economic results
• Procedure:
1. Connect to:
https://www.iit.comillas.edu/rnm/ReferenceNetworkModel.html
2. Upload an street map image.

Further work
2.1 The color of the buildings must be clearly different from the color
of the streets in the selected image.
2.2 The scale of the selected image must be close to 0.5pixels/meter.

2.3 Set the scale
Further work

2.4 Click in two points to set a distance in pixels
2.5 For the selected distance in pixels, set the correspondence in
meters, to obtain a scale in pixels/meter.
Further work

2.6 Select the building block colors
• All the colors inside building blocks have to be selected.
• Streets must not be selected.
Further work

3. Set the technical parameters of consumers
4. Check the parameters of the library of equipment
Further work

5. Build the network and save the output files.
Further work

C. Mateo Domingo, T. Gómez San Román, Á. Sanchez-Miralles, J. P. Peco Gonzalez, and A. Candela Martinez, “A Reference Network Model for Large-
Scale Distribution Planning With Automatic Street Map Generation,” IEEE Trans. Power Syst., vol. 26, no. 1, pp. 190–197, Feb. 2011.
J. A. Garcia Conejo, G. Tevar Bartolome, A. Gomez Exposito, and M. Rodriguez Montanes, “ANETO: A System for the Automatic Generation of
Theoretical Network Models,” 9th Int. Conf. Electr. Power Qual. Util., Oct. 2007.
M. B.-O. Larsson, “The Network Performance Assessment Model - Regulation with a Reference Network,” in Proceedings of Market Design, 2003.
T. Jamasb and M. Pollitt, “Reference models and incentive regulation of electricity distribution networks: An evaluation of Sweden’s Network
Performance Assessment Model (NPAM),” Energy Policy, vol. 36, p. 1788, 2008.
C. J. Wallnerstrom and L. Bertling, “Investigation of the robustness of the Swedish network performance assessment model,” Ieee Trans. Power Syst.,
vol. 23, pp. 773–780, May 2008.
A. Navarro and H. Rudnick, “Large-Scale Distribution Planning — Part I : Simultaneous Network and Transformer Optimization,” IEEE Trans. Power
Syst., vol. 24, no. 2, pp. 744–751, 2009.
T. Paulun, H. J. Haubrich, and C. Maurer, “Calculating the Efficiency of Electricity and Natural Gas Networks in Regulated Energy Markets,”2008 5th
Int. Conf. Eur. Electr. Mark. Vols 1 2, pp. 94–98, 2008.
G. Schweickardt and V. Miranda, “A two-stage planning and control model toward Economically Adapted Power Distribution Systems using analytical
hierarchy processes and fuzzy optimization,” Int. J. Electr. Power Energy Syst., vol. 31, pp. 277–284, Jul. 2009.
E. Garcia, G. Schweickardt, and A. Andreoni, “A new model to evaluate the dynamic adaptation of an electric distribution system,” Energy Econ., vol.
30, no. 4, pp. 1648–1658, 2008.
K. Kawahara, G. Strbac, and R. N. Allan, “Construction of representative networks considering investment scenarios based reference network
concept,” presented at the Power Systems Conference and Exposition, 2004. IEEE PES, 2004, p. 1489.
V. Levi, G. Strbac, and R. Allan, “Assessment of performance-driven investment strategies of distribution systems using reference networks,” Iee
Proc.-Gener. Transm. Distrib., vol. 152, pp. 1–10, Jan. 2005.
G. Strbac and R. N. Allan, “Performance regulation of distribution systems using reference networks,” Power Eng. J., vol. 15, pp. 295–303, Dec. 2001.
Virendra Shailesh Ajodhia, “Regulating beyond price: Integrated price-quality regulation for electricity distribution networks,” PhD, Delft University of
Technology, 2005.
J. W. M. Lima, J. C. C. Noronha, H. Arango, and P. E. S. dos Santos, “Distribution pricing based on yardstick regulation,” Ieee Trans. Power Syst., vol.
17, pp. 198–204, 2002.
“Benchmark Systems for Network Integration of Renewable and Distributed Energy Resources,” CIGRE WG C6.04, 2014.
Further readings

Q&A
Contacts:
Dr. Carlos Mateo
Email: Carlos.Mateo@comillas.edu
Dr. José Pablo Chaves Ávila
Email: Jose.Chaves@comillas.edu
Institute for Research in Technology
Comillas Pontifical University
Santa Cruz de Marcenado 26
28015 Madrid, Spain

Reference Network Models

Recommended

More Related Content

What's hot (18)

Similar to Reference Network Models (20)

More from Leonardo ENERGY (20)

Recently uploaded (20)

Reference Network Models