PAPER - ESTIMACIÓN DE LA DEMANDA ELÉCTRICA DIVERSIFICADA

UPS-LC001

1

Estimation of Diversified Electrical Demand for Energy Optimization in Residential Buildings Cevallos L., Student Member, IEEE

Abstract—often, one of institutional policy of utility companies is to provide an optimal and reliable electric service. This means that as a public or private institution in the power sector, these must comply with the provision of electric service in the best conditions of efficiency, continuity, quality and sustainability; in addition, the company must continuously promote technological innovation to provide a distribution of less environmental impact and more responsible on use of electricity. As a strategy, one of the main objectives of utility companies is deliver energy in the best conditions according quality standards, in order to generate the highest level of satisfaction in subscribers or consumers. Under this premise, diversified electrical demand is not being calculated according to current reality of a domestic electric demand, especially in cities of Latin America where construction of modern buildings is in constantly growth. The main purpose is to obtain real demand parameters, to estimate by calculation power demand as close as possible to a correct determination of the same. The goal is determination of electric demand, which should not be over or underestimated for a modern building project in an urban area of a city. Index Terms—Customer demand, Optimization, Utilities, Government, Buildings.

Measurement,

housing from the government. This growth demands the need for a power supply with updated design parameters that include factors for estimating chords with diversified demands of consumers, which is the subject of analysis in this paper.

II. GENERAL ASPECTS A. Electricity Demand Forecasts at 2021 According to the utility which provide electric service to the city, it has determined by statistical analysis that by 2021, the change in the pessimistic projection is less optimistic to the 3.76% for energy and 3.55% for the power in MW. These differences are considered reasonable at end of year forecast of energy demand and the system under normal power supply. Furthermore, these values have been the basis for forecasting electricity demand by substations. [5] ELECTRICITY DEMAND FORECASTS for2021

I. INTRODUCTION Is generally known that several Latin American cities and especially -those that are capitals of their countries-, are experiencing a constant growth in the area of building construction as a result of an economic development or due to a natural increase population according to their birth rates in the last decade. Such is the current case of the Quito city, capital of Ecuador, which coincidentally with the beginnings of a economic changes and particularly with the modify of its local currency for the U.S. dollar –dollarization-, this city is breaking through social and political transformations that keep the development of a growth economic model, particularly in the area of construction industry, and as a result, there is an increase in energy demand of consumers , and that, according to data government entities such as the Central Bank and the Chamber of Construction , determine that the growth in construction is 27 % compared to its GDP . Indeed, changes, development and transformation of the city have been caused by the construction of civil works and residential buildings, condominiums, residential, commercial and industrial contractors, this real estate development programs implementation constructs are added large scale

PRONÓSTICOS DE DEMANDA ELÉCTRICA AL 2021 PARA EL SISTEMA ELÉCTRICO QUITO SEQ [MW] Year

2016

2017

2018

2019

2020

2021

Optimist

838.7

875.8

914.5

954.9

997.1

1041.1

Probable

830.2

865.4

902.1

940.3

980.2

1021.8

Pesimist

822.3

855.9

890.8

927.1

964.9

1004.2

Table 1. ELECTRICITY DEMAND FORECASTS FOR2021 Source: EEQ S.A. Plan de Expansión de Demanda Eléctrica del SEQ, Cap. 3.

The growth of current demand and its future projection as shown the above table reveals that with the most pessimistic option of future demand to 2016 will increase by approximately 1 %, which means megawatts an increase of 122 [MW]. In this pessimistic scenario, the ratio for the year 2021 the demand will increase 30 % over 2012.

UPS-LC001

2

According to the expansion plan of the Utility, there is no dates projected for 2021 in residential power consumption, only shows that by 2010 the turnover became 39.7 % of the total consumption equals 1.29 GW. But on the contrary, have projected data issued by another governmental entity that is the National Electricity Council who said that the turnover of watts / hour [Wh ] reach the average figure of 11.86 Giga Watt hours representing 40 % of projected total billing . With this information it is confirmed that the area of residential consumption accounts for 40 % share of total electricity demand, as well as detailing the table shown below

6,264 6,667 7,039 7,593 7,954 8,333 8,712 9,120 9,564 32,0%

Street lighting & Other 2,196 2,272 2,345 2,417 2,489 2,564 2,640 2,717 2,794 9,0%

17,614 18,676 19,686 21,336 23,113 24,862 26,604 28,201 29,629 100,0

7,9%

3,5%

7,5%

Year

Residential

Comercial

Industrial

2012 2013 2014 2015 2016 2017 2018 2019 2020

5,742 6,065 6,381 7,166 8,268 9,317 10,355 11,213 11,860 40,0%

3,412 3,672 3,921 4,160 4,402 4,648 4,897 5,151 5,411 18,0%

Growth 2009-2020

8,8%

6,7%

Participation 2020

total

Table 2. Invoiced Consumption Projection (GWh) Source: Conelec. (2009). Plan maestro de electrificación, Demanda Eléctrica, Cap. 5

It is very important to note that the regulator entity of the energy belonging to the government, has made estimates taking into account the historical statistical data billing analyzed time series and adding the element of change of the energy matrix, it means, the replacement of consumption liquefied gas used in kitchens for the consumption of electricity since 2016. B. Projecting the Electrical Service Users Utility expansion plan by 2021 has not determined the projections of the number of subscribers that probably exist at that time, but as confirmed by 2010 data subscribers were 724,447 residential sectors. Meanwhile the regulator entity has developed projections in this area at the national level, obtaining an estimated factor of subscribers to

the Electric System of the city. According to the governing entity for the year 2010 has registered in the residential sector 3'378 .435 subscribers nationwide, while the Utility recorded 724,447 subscribers in the electrical system of the city. The relationship between these two figures allows a factor of 21.44 % which serves as a parameter estimation 2020. Therefore, if the regulator estimated for 2020 4.322.718 subscribers multiplied by the factor given above, we can estimate that the number of subscribers in the electrical system of the city shall be 926,932. C. Consumer Increase Projection A particular analysis done by the government regulator, some factors were taken for projection of increase subscribers by 2020 , which says the following: "As for the perspective of consumers or users of electricity, considerations analyzes have taken into account the projection of future results would the average annual consumption per unit of main areas of consumption. In this sense , according to the increase in the efficiency of electrical appliances as well as their use, but also incorporating cooking and electric water heaters are expected in the plan period, an average increase in usage in the by 9% per year for the residential sector; -1% to -3.1 % for commercial and industrial . In particular, lighting and residential use, is incorporating the use of lamps (bulbs) compact fluorescent type, thus consumers compensates to some extent increases consumption by cooking, showers and water heaters. Within the national grid, the average annual growth of consumers is expected to be 2.6% in the lower growth scenario, and 3.1% and 3.5 % respectively for the scenarios medium and higher growth in the period 2008 -2020. These growth rates take into account the recovery of irregular consumers, as part of plans to reduce electric losses “[6]. III. CATEGORÍAS DE MODELOS DE ENERGÍA Currently there are several models that support for analysis of energy systems and energy planning that can be developed for a short or long term.

UPS-LC001

Moreover these analyzes can also study the impact of energy policies. The models are based on different approaches where mathematical tools that allow a classification according to different factors are used, for example: • Classification according to purpose / objective of the model • Identification by spatial coverage • Identification by Classification modeling approach "bottom up" versus "top down" A. According Purpose / Objective Model Using the criterion of purpose / objective energy models can be classified into the general categories of Demand Models, Offer Models and Designs Systems. The Demand models have a main function of their prognosis. In offer models the goal is the prediction and planning of supply and system models are used to analyze the energy system as a whole, including supply and demand. B. Classification According to Space Coverage Generally, energy models are developed for purposes of national planning or analysis of global politics and therefore a first classification considers national and global models, although there are also regional. Electricity demand estimates for much finer spatial resolutions that can be defined at the node level or through grids that are usually hexagonal. C. Classification According Modeling Approach The main approaches of models used for policy evaluation, planning and implementation of energy systems forecasts are: • Optimization • Simulation Models and Partial Equilibrium • Use models or Final Accounting • Econometric Models

3

• Computable General Equilibrium (CGE) • Hybrid Models [4]. IV.

DIVERSIFIED DEMAND ESTIMATION

A. Maximum Diversified Demand Curve Utilities should possess an optimal design which supports maintaining quality standards and be at the forefront of the growth potential in a sustainable manner for the Company. The overall purpose of distribution system planning is to minimize the operating cost of substations, priced power cables, distribution transformers, secondary networks, power losses and energy, subject to restrictions of allowable values of voltage, momentary voltage drops and continuity of service. For projection, design, construction, operation and maintenance of a distribution system must meet requirements as the application of the rules of electrical codes, cost optimization, location of power to the system , knowledge of the loads, knowledge of growth rates of charges, information on the location of charges for residential, commercial and industrial areas , socioeconomic status, and location of different facilities in the area, selection of distribution equipment including switches , lines, transformers and conductors , and others. These studies provide lift diversified maximum demand curves expressing the same behavior of consumers connected to a distribution system. Failure to develop and have adequate curves reflect the behavior of a system can lead to significant electrical company operating problems such as overloads in the feeders when maximum capacity is exceeded, overloading of transformers when its power capacity is exceeded and oversizing same if consumption is too less than the installed load, producing this increased losses of transformers, plus it is expressed by not getting the economic optimization of the costs of equipment.

UPS-LC001

4

For certain settings to the exponential loading curve equation shown in the following expression is used:

=

+ c

is the value of the dependent variable "y" when the independent variable " x " has a value of 0 . The value of “b” defines the slope of the curve and the constant “c“represents the horizontal abscissa, which stabilizes the curve and which becomes a constant value. To obtain the parameters of the variables a, b, c and start setting points should eliminate outliers that are within the graph of the curve as this will impair the performance of the exponential computation. Figure 1. Tipical curve of demand

Some utilities of distribution and trading of power settings load curves of other regions adapted to your system. This technique is based on observing and comparing the behavior of residential users fractionating by socioeconomic strata, obtaining similarities in consumption between the different areas, avoiding costs in the study design and measurement. However, it should be noted that the region population have social, cultural, climatic and /or customs differences, and to make the adjustments and settings from one system to another, the designs of a demand draft inconsistencies may have either short or medium term, which would be reflected in the estimation of overloaded , oversized equipment or sub estimated, plus it can present operational problems in feeders and transformers, inadequate levels of stress, increased technical losses and hence increased economic losses.

Figure 2. Dates of demand (loads)

On the other hand, a curve setting need to establish a correlation coefficient which will determine whether there is relationship between the 2 variables, otherwise run the regression to define the line that best fits the set of points.

With the aforementioned is imperative that each utility perform appropriate studies to obtain a certainty in the results of the maximum diversified demand curves and this permits the behavior of the differentiated, segmented by socioeconomic residential demand and various user groups. B. Curve Adjustment Method

Figure 3. Curve of Variables Correlation

UPS-LC001

5

It is usually used the MatLab program which allows adjustment of the curves with several alternatives according to schedule [9].

[3] Stetson, L.E. ; Nebraska Univ., Lincoln, NE, USA ; Stark, G.L.Peak electrical demands of individuals and groups of rural residential customers. http://ieeexplore.ieee.org/xpl/articleDetails.jsp?tp=&arnu mber=8978&queryText%3DResidential+electrical+dema nd [4] Comisión Nacional de Energía. Chile. Demanda Energética Nacional de Largo Plazo. Modelo de Proyección.2009 [5] EEQ S.A. Plan de Expansión de Demanda Eléctrica del SEQ, Cap. 3. Pronósticos de Demanda del SEQ al 2021

Figure 4. MatLab Fitted Curve

From Figure 4 it can be deduced that the exponential fit (green curve) is fit best modeling of the points with a negative exponential curve to the point where it stabilizes and becomes a constant value. V. CONCLUSIONS

In a changing and growing territory as according its power demand, estimating it requires the cooperation of important factors such as the actual consumption in watts of users and the number of subscribers that make up a building. This requires planning and land identified by sampling the best sector for development of the estimated diversified demand. Moreover utilities once on study data and results should be part of their statistical and thus implement a management system consumption data subscribers and diversified peak demands and daily schedules peaks, all in order to improve, enhance and put to good use of power distribution.

REFERENCES

[1] Urban, G. J. ; Vermeulen, H. J. The statistical Modelling of Residential Electrical Demand for the Evaluation of Impacts that Result from Demand Side Management Interventions. http://ieeexplore.ieee.org/xpl/articleDetails.jsp?tp=&arnu mber=6125544&queryText%3Delectrical+demand [2] Marinescu, A. ; Sch. of Comput. Sci. & Stat., Distrib. Syst. Group, Trinity Coll. Dublin, Dublin, Ireland ; Harris, C. ; Dusparic, I. ; Clarke, S. Residential electrical demand forecasting in very small scale: An evaluation of forecasting methods. http://ieeexplore.ieee.org/xpl/articleDetails.jsp?tp=&arnu mber=6596108&queryText%3Delectrical+demand

[6] Marcelo Meira. Medardo Caden. Gina Moreta y Otros., “Plan Maestro de electrificación CONELEC 2009-2020.” Quito, pp. 1–506, 2009. [7] Electrical Transmision and distribution Westinghouse Capitulo 2 Método de Arvidson..Redes de distribución de energía pagina 37, publicación titulada “Diversified demand method of estimating residential distribution transformer loads” [8] Proyecciones de demanda de energía eléctrica y potencia máxima Demanda, proyección de demanda 2003-2011 Online: www.siel.gov.co [9] Rosario Ruiz Baños, Departamento de Biblioteconometria y Documentación. Universidad de granada (España). Modelos No Lineales.