MILETO was
developed by UNION FENOSA in order
to determine an Optimal Reference Network that could reflect the reality
of the market with a reasonable degree of accuracy and would therefore
produce quantified costs acceptable to the Distribution Companies.
This was a project for which engineering and development were executed
by SOLUZIONA engineering.
Scope
 The
Network Model to be developed should identify and model the market
rigorously, using: the physical units minimizing investment and losses,
the financial assessment of Fixed Assets, and the annual value, in
order to calculate the Preventive and Corrective Maintenance costs,
determine and assess the Losses, determine and assess the Undistributed
Energy, and to consider the Quality and Reliability that guarantee
the completion of the Indices(TIEPI and NIEPI) by zones.
Furthermore, the Model should be compatible with the inventory from
which it was taken (from the Electric Substations, or SSEE, of Distribution
to the Injection Points of the Transport Network). An important requirement
of the Model is for it to take into account the significant differential
aspects, both technical and geographical, because of their influence
on the model and the cost.
Development
To achieve this objective SOLUZIONA engineering
has made the necessary engineering analyses and developments,
which have been converted into an information tool: the Income-Producing
Electric Distribution Model called MILETO which models a reference
network for Low Voltage users up to High Voltage injection points.
MILETO models the network that connects and distributes energy to
each and every one of them.
Entry data
The model's lower limit is based on each of the geographical coordinates
and the data on contracted Capacity and annual energy consumption
of all users of low, average and high voltage. The model's higher
limit is set by the coordinates of the High Voltage injection points,
that is, all or some of the Substations of the Transmission Network
with usable capacity and necessary for Distribution.
Entry data include the library of components and installations with
their assessment of materials and manpower. This is a modular library
which allows enabling or disabling installations, adding and reforming.
Entry data also include the Maintenance Ranges, Failure Rates, repair
Times, References and service Quality indices by Zones, municipal
and provincial borders, geographical elements, etc.
When the coordinates and data on the Distribution Substations and
Transformation Centers are available, one can compare the modeled
network with the actual one.
Objective
 The
objective is to establish the best possible reference frame for determining
the Distribution Rate.
We can summarize the possible approaches into three:
1.- IPC-X/kWh Rate,
2.- Comparison between companies
3.- Reference network model.
When the approach is based on applying a IPC-X/kWh distributed rate,
adjusting the regulator -- the efficiency factor X which is inferred
from the companies' financial results, a global and general approach
is being made. This does not reflect geographic nor market reality,
nor the resulting network needed to maintain supplies, by legal obligation,
to complete all the zonal quality indices that are to be defined.
To establish a precise comparison between companies, it is necessary
that these be similar. In Spain they are not similar, neither in terms
of size nor in homogeneity of the market. The third approach, which
we understand to be more adequate, is to build a reference Network
Model adapted to the geographical and market reality.
What should be the scope of such a Model? There are three options:
- A scope that completely models the entire Network.
- A scope which is based solely on inventory.
- A scope that combines a partial network model and inventory.
If we had a complete inventory of all the Distribution installations
in their different voltage levels we would not need a model. It would
not even be necessary to send signals on the rating since the theoretical
rating that would be given by the real network could be calculated
and gauged using the declared incidences.
The real situation is quite different, as there is no complete nor
reliable inventory in BT, as there is in MT and AT where it is possible,
because it deals with less volume. Producing a new and complete inventory
is deemed to be a very arduous task that would not make sense. For
example, gauging and auditing the RBT Subway would have a very high
cost.
It is a task which the companies can undertake to gain a more thorough
knowledge of their networks and to facilitate maintenance, operation
etc.
 In
a complete scope Optimal Reference Network the reliability of the
same can be questioned because of being far from the reality it attempts
to replicate.
A semblance may be required in the order of magnitude of the modeled
and real installations. For example the number of CCTT, of machines
and the Power installed.
Let us remember that this is not about making a stylish nor academic
exercise in which the volume of installations may be indifferent,
but of compensating the costs derived from existing installations
with a credible model adjusted using adequate signals by rating that
allow the establishment of bonuses and penalties.
The measuring of the BT, and of each Transformation Center plus its
RBT, is reasonably stable in time and optimal given its reduced size.
Thus, a model can reproduce this situation with a good semblance to
reality.
In MT the situation is quite different. The situation of each of the
Substations -or SSEE- that exist is a result of reconciling available
spaces, political administrative authorizations of entities from different
spheres, technical needs, geographical arrangements, etc.
It could be assumed that at the time of its planning its position
would be optimal; as time passes, with the load evolution, neither
the number nor the location nor the individual capacity of each SE
will be similar if we use these in the model.
If we model several years consecutively we will find movable SSEE
which do not allow valid financial signals to be produced.
We understand that to compare a Model against the Inventory is not
necessary since it is possible to reconcile both approaches.
Our proposal is to adopt a partial scope Model patterning the BT and
the MT feeders to connect the Inventory Substations even at its own
voltage level and starting from there with the actual network.
This presentation is algorithmically more complex than complete modeling
since the model's degrees of freedom are reduced. However, a solution
is achieved that is optimal for some given positions of the SSEE,
guaranteeing the convergence with reality and allowing rating signals
to be actually contrasted.
Setting
The essential aspect to consider in order to have a clear idea of
the magnitude of the challenge is to process each and every one of
the data of the BT, MT and AT users -- more than 20 million coordinates.
In addition, the values of the contracted power are different for
each user, being discreet values, and their distribution is given
by their real position which require aspects like leakage or concentration
to be treated very seriously.
 Identification
and Modeling of the market
With the individual information of the X,Y,Z UTM coordinates of each
user the population cores are identified with parameter criteria of
absolute or relative geographical proximity (density). The town structure
is automatically modeled, and the actual streets and blocks are pointed
out resulting in a town model similar to reality.
 The
structured towns generate a model defined by streets and blocks (closed
premises) and the minor or not so well-defined cores appear as polylines
(open premises) which join, as in reality, the current consumption
with minimum distance trees.
This is the model most similar to the actual current consumption since
it is generated from them, and it is updated, doing away with errors
due to homes without electricity, etc. It is also the most economical
since it does not require additional information..
Alternatively the MILETO Model can work against reference maps, digitized
or cadastral mappings. This alternative can be more accurate but is
much more costly and presents some problems that are difficult to
solve such as the different updating degree of the mapping against
the homogeneity of the users' coordinates, the lack of mapping for
a big number of population cores, its elaboration with various criteria
that require dealing with an information systems which is slow, complex
and that produces uncertain results.
Sector Types
One could define sector types of a territory, conduct an in-depth
analysis, build the adequate optimal network with planning criteria
and later extrapolate to the rest of the territory's sectors.
But considering the need for complete identification and modeling
of the market as proposed by MILETO, this alternative presents a clear
disconnection with reality, as it identifies characteristics and extrapolates
them to supposedly similar sectors.
This would the produce an illicit generalization because markets are
complex social, financial and geographical entities that cannot be
reduced or simplified to a few model sectors.
Categorization
The population cores to be given electricity are to be categorized
based on their size (clients, loads) as urban, rural and remote. This
categorization, while automatic, is one of the most delicate aspects
of the process as it dictates the mode of electrification.
Thus, for example, to categorize a core as remote implies utilizing
the Tensioned low voltage network, Transformation Centers -CCTT- concrete
transmission lines and in addition being optimized with other cores
so as not to designate a CT a priori in each core such as occurs in
reality.
To categorize as rural or urban also has consequences on the use of
certain types of installations, even when within each group they may
be inhabited or uninhabited (leaving at least one model) to have a
flexible configuration.
It would appear that remote cores would be excluded from Underground
installations, but for the CCTT houses or buildings will appear, as
well as the laid network, the underground network, the mesh architecture,
etc.
MILETO minimizes the consequences of this aprioristic rating since
it supports an overlapping of types of installations which can be
used in the cores located in the remote, rural and urban separation
limits.
|
you're here: home
/ resources / publications
/ technical articles / mileto: a reference network
model 
