Design and Inteligent Control of Hybrid Power System in Telecommunication

Design and Inteligent Control of Hybrid Power System in Telecommunication Boban Panajotovic and Borislav Odadzic Republic Telecommunication Agency, Republic of Serbia Belgrade, Visnjiceva 8, Serbia Email: boban.panajotovic@ratel.rs; borislav.odadzic@ratel.rs Abstract - The basic requirements for telecommunication power systems are related to their safety, long life and uninterruptible power (voltage). Hybrid power system design for power feeding of telecommunication equipment has to fulfill same requirements. The subject of this paper is hybrid power system requirements, design, practical calculation and sizing of its major part: photovoltaic cell, wind turbine, diesel generators and storage battery. Focus of this paper is inteligent control based on weather condition and battery status in this kind of system and benefites as: fuel consumption and CO 2 emmision reduction and logistic cost decreasing. I. INTRODUCTION The European Commission today called on Europe's information and communication technologies (ICT) industry to outline by 2011 the practical steps it will take to become 20% more energy efficient by 2015. ICT equipment and services alone account for about 8% of electrical power used in the EU and about 2% of carbon dioxide emissions. But using ICT in a smart way could help reducing energy consumption in energy-hungry sectors such as buildings, transport and logistics [1]. Due to new power and energy context such as greenhouse effect and other environmental issues, fuel depletion and electricity cost increase, new regulation and standards, telecom operators have to make efforts for using renewable energy solution [2]. In telecommunication the renewable energy sources, because of the high cost for Wh, are generally used in remote areas where the public mains is unavailable [2]. The need for renewable energy may also enable telecommunication services (areas with no power grid), to expand coverage and to deploy high data rate services (active equipment in the access network) [2]. In nowdays telecom operators and Governments have to make efforts for using renewable energy solution wherever is possible, like in big telecom and date centers, not only in remote areas where the public mains is unavailable. Action plan of Serbian Government is to stimulate the use of renewable energy sources and developing of these technologies as the future chance for Serbian industry and society. In telecom area this plan is managed by Republic Telecommunication Agency, which tries to promote and stimulate the use of renewable energy sources by telecom operator. Such renewable energy sources are: Fuel Cells, Photovoltaic cell, Wind Turbine Generators, Micro hydro Generators, Stirling machine, renewable cooling sources like geo-cooling or fresh air cooling, etc. In telecom application an efficient and reliable solution is to combine renewable and traditional energy sources. Hybrid power system capture the best features of each energy resource and can provide grid-quality electricity. Achieving higher reliability can be accomplished with redundant technologies and/or energy storage system. Some hybrid systems typically include both, which can simultaneously improve the quality and availability of power which are major requirements in modern telecommunication [2] (goal is to achive uninterruptible power feeding of telecommunication equipment with quality power). Once the system is installed and started up, it is necessary to keep the system at its best performance level. The cites reliability studies play an important role [5]. This work will present requirements, design, calculation and sizing with focus on inteligent controling of hybrid system, which is combined from the next major part: Photovoltaic cell, Wind Turbine Generators, Diesel (or gas) back-up generators and storage battery. In this work, calculations will be done practically, with real data, parameter, requirements and component sizing, and present practical guideline for hybrid power system design and calculation. Focus of this work is on design, calculation and sizing of hybrid power system for use in telecommunication, but principles and methods can be used in other applications. II. REQUIREMENTS The basic prerequisites imposed to telecommunication power systems are related to their safety, long life and uninterruptible power [6,7]. Hybrid power system design for power feeding of telecommunication equipment has to provide quality uninterruptible voltage (AC, DC or both). Due to the renewable energies intermittent behaviour, this kind of system has to be oversized [5]. In order to avoid very high costs, an optimisation method should be used and good one is described in [8]. One of the major requirements is to design hybrid system cost effectly, with minimum exploatation costs. 978-1-4244-5794-6/10/$26.00 2010 IEEE 1453

A. Photovoltaic cell Sunlight may be used to generate electricity directly via photovoltaic cells. The problem is low efficiency of photovoltaic cell which can present problem in situation where there is not enough space for required cell mounting. Problem is still high price per Watt of photovoltaic cells. One of the major requirement in hybrid power system design is to optimize size of cell. B. Wind Turbines The output power depends of wind velocity. If the wind speed changes very smoothly, the output power of wind turbine will also change very smoothly. On the other hand, wind turbulence causes the output power to fluctuate [3]. Requirements for wind turbines is to give constant output power as much as it possible in wide range of wind velocity. Benefit for using is competitive price of wind turbines. C. Energy Storage There are different types of energy storage (flywheel, battery, fuel cell etc.). The energy storage behaves like a large buffer to accommodate the unequal instantaneous energy in power system [3]. In hybrid power system which is subject of this work, battery used like energy storage system work in daily charging and discharging. Basic requirement for this battery is high number of cycling. Opposite, replacing of battery can present significant exploatation costs, which is not environment friendly. D. Diesel (or gas) back-up generators Conventional back-up generators are normally diesel engine directly coupled to synchronous generators. In hybrid power system which is subject of this work, diesel (or gas) back-up generator has to ensure uninteruptibility of the system. One of the requirements is to optimize size of generator and to minimize fuel (gas) consumption and carbon dioxide emissions. Diesel generators must be able to keep the power balanced when the wind turbine or load varies [3]. III. SYSTEM DESIGN There are many developed methods and algorithms for Hybrid power system modeling, design and components sizing. For optimal modelind, design, calculation and components sizing is essential to have precise micro location meteo data and to define system reliability in power feeding. A detailed study of the above factors is the first step to choose the required system topology and to make the best of the local potentialities to supply the telecommunication equipment [5]. For example, hybrid power system can be modelled with a high degree of accuracy considering the higly complex working of actual systems. Optimisation algorithms change the values of decision variables of an underlying in model in such a way as to optimise the resulting value of the model s objective function [4]. In this work calculation and sizing is done for next basic required parameters: -location: central-eastern Europe (Serbia, Belgrade), -meteo data: real 4 seasons (summer, autumn, winter and spring) -temperature range: -15 ºC to +40 ºC, -real problem: in period from November to March can occur heavy fog days which can last for more than 7 days (period of black out). Negative in this situation is absence of wind, -period of insolation in Belgrade per year: ~2400h, -average period of insolation per day: ~7h, -average insolation per day during summer season:~9h, -nominal voltage: 48VDC, -range of working voltage (input voltage of telecommunication equipment): 40,5VDC to 57VDC, -DC load (telecommunication equipment): 1000W, -DC load current (for calculation): 20A, -required autonomy of Diesel generator back-up (need for season fuel tank sizing): 15 days constantly, -required autonomy of battery back-up (need for battery capacity sizing): 2 days constantly. Note: This requirement has big influence on hybrid power system price. In region with constantly sun insolation during all year (Africa, Greek islands, Spain see-coast ), battery back-up can be determined for 24 hours. In region with no constant meteo condition during all year, longer autonomy higher capacity is recommended. It requires higher battery charging current. If diesel generator back-up fuel consumption has to be minimized, energy for battery charging has to be provided from photovoltaic cell which present most expensive component in the system. The energy for battery charging can also be provided from wind generator, but in location where system will be installed wind potential is not constant. This is the reason why meteo data for micro location has to be precise. Correct data is necessarily for system optimizing. IV. CALCULATION AND COMPONENT SIZEING In this work calculation and component sizing are based on DC current (20A) and 48VDC nominal voltage. Converters efficiency will not be taken in consideration, because it is not essentially for this work. Figure 1 shows hybrid power system described in this work. A. Capacity of battery Required minimal capacity of battery can be calculated according formula: Q min = (I DC x Т x к 1 ) (Аh) where I DC is telecommunication equipment DC current (20A), T is required authonomy (2days = 48h), к 1 is coeffitient for 1454

battery capacity increasing depend of plate sulfatisation and low temperature (1,15). Q min = 20A x 48h x 1,15 = 1104 (Аh) In this system, required autonomy will be ensured with two 600Ah, 48V batteries in GEL technology. Q = 2 x 600 (Ah) > Q min = 1104 (Аh) B. Current for battery charging Battery manufacturer usually recommend maximum 0,10 C 10 battery current for battery charging. Some manufacturers, for their battery allowed 0,15 C 10 battery current for battery charging. C 10 = 2 x 600 (Ah) = 1200Ah I batt =0,10 C 10 = 120A I battmax = 0,15 C 10 = 180A (maximum battery charging current). C. Area of photovoltaic cell Many authors developed different methods and algorithm for optimizing of photovoltaic cell. In hybrid power system components sizing, photovoltaic cell present critical point because of its price. Principe is to provide energy from photovoltaic cell for load feeding and battery charging. In this work requirements for photovoltaic cell are defined for summer period. Sizing will be done according to this requirements. Day (24h) will be split in two period. First period, with maximum insolation of 9h (in this period photovoltaic cell provide maximum power), second period of 15h (totally black out, no energy from photovoltaic cell). Requirements for photovoltaic cell in first period are to provide energy for telecommunication equipment power feeding and energy for battery charging. In this period (insolation of 9h) battery must be charged with enough energy to provide back-up autonomy in situations for 2 days with no sun (not only for 15h black out). For micro location which is the subject of this work, this is a typical situation. Energy can be provided from diesel generators also, but the target is to reduce fuel consumption of diesel generators. Bigger area of photovoltaic cell means more energy during other season also, not only summer. Final results are total operation cost and CO 2 emission decreasing. I method In this method, calculation will determine energy from photovoltaic cell in period of 9 hours for telecommunication equipment power feeding and for energy storage in battery. E PV = I DC x U DC x T ins + I DC x U DC x T back-up where I DC is telecommunication equipment DC current (20A), U DC is nominal voltage (48V), T ins is period of insolation (9h) and T back-up is required time for battery backup (48h). E PV = 20A x 48V x 9h + 20A x 48V x 48h = 54,720kWh Required power from photovoltaic cell in period of insolation is: P PV1 = 54,720kWh / 9h = 6,080kW Area of photovoltaic cell In this work we adopt photovoltaic cell with maximum power (P m 2 ) of 135W/m 2. S PV1 = P / P m 2 = 6080W / 135W/m 2 = 45m 2 II method In this method, calculation will determine power from photovoltaic cell required for telecommunication equipment power feeding and for battery charging with maximum current (0,15 C 10 ). Calculation can be done also for 0,10 C 10 battery charging current, and derived value of photovoltaic cell area of 50m 2 also fulfill hybrid power system requirements. P PV2 = I DC x U DC + I battmax x U DC where I DC is telecommunication equipment DC current (20A), U DC is nominal voltage (48V) and I battmax is maximum battery charging current (for battery in this work I battmax = 0,15 C 10 = 180A). P PV2 = 20A x 48V + 180A x 48V = 9,6kW Area of photovoltaic cell As explained, in this work we adopt photovoltaic cell with maximum power (P 2 m ) of 135W/m 2. S PV2 = P / P m 2 = 9600W / 135W/m 2 = 71m 2 Conclusion for photovoltaic cell area According to previous calculation where S PV2 > S PV1, both calculated areas of photovoltaic cell fulfill defined requirements. Decision for adopting is based on many different facts as price, real meteo micrlocation condition, operational costs, etc. In this work authors made decisions to choose photovoltaic cell area S PV2 of 71m 2. Reason for this is request for reduction of diesel generator fuel consumption and CO 2 emission, decreasing of logistic costs and to present inteligent control in hybrid power system, which is not possible in system with area S PV1 of photovoltaic cell. Note: Smaller area of photovoltaic cell can be applied. For the region where the system will be installed and for power consumption of 1000W, photovoltaic cell of 20m 2 can fulfill some requirements. Problem in the system with smaller area 1455

of photovoltaic cell is the increase of fuel consumption and CO 2 emission. Also, logistic operational cost increase, more visit of location for diesel generator tank filling and services, more training person needed, etc. In the future, technology development will bring more efficient photovoltaic cell, and cell price per W will go down, final result will be same power from smaller area, for less money. D. Power of wind turbine Power of wind turbine can be determined at different ways. Sometime, wind turbine is sized only for power feeding of equipment (load). In hybrid power system with solid area of photovoltaic cell, price of wind turbine is minor compared with the price of the other system component. Basic parametar is power of wind turbine. Other parameters and working characteristics depend mostly of microlocation wind potential, wind velocity, etc. In this work only power of wind turbine will be determined, the other parameters and wind turbine working characteristics are not esential for this work. Requirements for wind turbine in this work is to feed telecommunication equipment and to charge battery with maximum current (for battery in this work I battmax = 0,15 C 10 = 180A). Power of wind turbine will be determined according to power P PV2 = 9,6kW. Output power of wind turbine depends of wind velocity. Based on wind potential in micro location which is the subject of this work, authors recommend wind turbine with power of 12kW. Note: oversizing of wind turbine can cause some operation problems. This problems and wind turbine regulation is not under the scope of this work. In this work minimum power of diesel back-up generator S min = P PV2 / cosφ = 9,6kW / 0,8 = 12 kva Authors recommendation is to choose bigger diesel backup generator, not according to calculated power. Recommend over sizing is approximately 20%. Some of the reasons are: more stable voltage and frequency, particularly during transitional period, possibility to add new equipment (load), etc. In this work authors select diesel back-up generators of 15kVA (cosφ = 0,8) ; 12kW, with electronic speed regulator. S = 15kVA (cosφ = 0,8) Sizing of diesel back-up generators fuel tank Diesel back-up generators of this power is usually equipped with daily tank from 50 to 100 liters. Daily tank is insufficient for application with no grid connection. Without additional (season) tank, in some periods during winter, it can cause operational and logistic costs to increase and reliability to decrease. Required autonomy for diesel generator back-up in this work is 15 days constantly (T aut ). Average consumption of diesel engine is 220g per kw per hour. Capacity of season tank for applied diesel back-up generators (1 liter ~ 1kg): V = 0,220l x P PV3 x T aut V = 0,220l x 9,6kW x 15days x 24h = 760 liters E. Diesel (or gas) back-up generators In this work, diesel back-up generators will be used for meteo condition without sun and without wind, when the back-up battery is discharged,. Positive fact is that already, at many location diesel backup generators are installed, or they are already in plan for installation at future location. General authors recommendation is to use diesel back-up generators bigger then 10kVA (cosφ = 0,8), in application for power feeding of telecommunication equipment or the other critical application. Smaller diesel back-up generators have not quality speed and voltage regulator, it cause unstable voltage and frequency. Also there are no possibilities to use sophisticate system for diesel back-up generators control and monitoring. Power of diesel back-up generator will be determined according power P PV3 = 9,6kW. S = P / cosφ Figure 1. Hybrid power system V. PRINCIPLE OF HYBRID POWER SYSTEM OPERATION Principle of hybrid power system operation is to provide energy from photovoltaic cell and wind generators for load power feeding and battery charging wherever is possible. If energy from sun or wind is not sufficient for battery charging, only load will be fed by renewable energy. If renewable energy source is not sufficient for load power feeding, difference of energy will be provide from battery, or from diesel back-up generator if battery is discharged. 1456

When diesel back-up generator fed load with power, it means that battery is discharged. In this situation diesel backup generator is used for load power feeding and battery charging. Battery in hybrid power system must be protect from over discharge by low voltage disconnection (LVD). The rating of low voltage disconnection is set as fix value (typically 1,8 V/cell, or 1,75 V/cell). Also, maximum battery charging current is set as fix value (typically I battmax =0,10 C 10 ). According algorithm described in this chapter most hybrid power system composed from photovoltaic cell, wind generators, diesel back-up generators and battery operate. VI. CONTROL OF HYBRID POWER SYSTEM Operation of hybrid power system is manage by controller. Logic and system control is very basic. Principles of system control is based on algorithm described in previous chapter. If energy from renewable sources is not sufficient, energy is provided from battery or diesel back-up generator. First back-up source is battery, then diesel generator. Because of that, battery has to be charged quickly as possible. Period of battery charging is longer if charging current is smaller. Opposite, period of battery charging is shorter if charging current is higher, but this can cause reduction of battery lifetime. In typical hybrid power system real battery status (battery condition) is not take in consideration. In this chapter, authors will explain inteligent control in hybrid power system, based on battery status (battery condition) and real meteo condition where system is installed. Parameters important for battery operation as low voltage disconnection and maximum battery charging current are usually set on fix value. In this work, those parameters will be programmable (changeable). For this approach, two requirements have to be fulfilled. Battery has to be monitored. Monitoring can be done for total battery, or for each cell. Each cell monitoring provides more precise data of battery capacity. Two capacity has to be calculated based on monitor information. Actual capacity (C a present instantly capacity of battery) and capacity of fully charged battery (percentage of nominal capacity). Battery can be safely used until capacity of fully charged battery is bigger than 80% of nominal capacity (C n ). If capacity of fully charged battery is nearby nominal capacity, higher charging current and lower level of LVD allowed. Weather parameters as air temperature, relative humidity, barometric pressure, wind velocity have to be measured, at present micro location. Nowdays technology enable, based on this value, to have very precise short term weather forecast (forecast). With measurement of this parameters, for particular micro location, 10 days weather forecast is accurate too. Based on battery capacity and weather forecast for particular micro location intelligent control philosophy will be explain in next examples. Example 1 Basic set of battery parameters: I battmax = 0,1 C 10 ; U LVD per cell = 1,80 V/ cell; Photo voltaic cell is design to feed load and charge battery with maximum current 0,15 C 10 ; Fully charged battery capacity = 98% C n ; Conclusion: battery is in good condition; Actual capacity C a = 30 % of C n ; Season: summer; Average insolation in this period = 9h; Short term forecast for one day: insolation = 7h; Forecast for next 3 days: heavy rain, no wind no insolation; After those 3 days, normal weather for summer season; In normal operation with fix battery parameters, in period of 7 hours insolation, battery will be charged with current of 0,1 C 10. After 7 hours of charging with this parameters, battery will be charge approximately at 70% of C n. With 70% of nominal capacity, battery can assure maximum 35-40h of autonomy. Nearly 35h of black out have to be covered with energy from diesel back-up generator. In this period battery will be charge from diesel generator. Efficiency of battery charging AC/DC convertor is approximately 90%. Convertor losses has to be covered by diesel generator. In method proposed in this work, intelligent control, based on battery status information and weather parameters, has to reconfigure next parameters: I battmax = 0,15 C 10, U LVD per cell = 1,70 V/ cell. If fully charged battery capacity is less than 85% C n (battery is not in good condition) these changes will not be enable. With maximum battery charging current of 0,15 C 10, battery will be fully charge for 7 h of insolation. It means autonomy of 48h (for U LVD per cell = 1,80 V/ cell). LVD will be changed at U LVD per cell = 1,70 V/ cell. This is approximately 6h more autonomy. With those changed parameters autonomy is 54 h. According this, 18h have to be covered with energy from diesel back-up generator. This is 17h less working time of diesel back-up generator, compared with system with fixed parameters (I battmax = 0,1 C 10, U LVD per cell = 1,80 V/ cell). Example 2 Season: winter; Average insolation in this period = 5,5h; Actual capacity C a = 30 % of C n ; Forecast for next 10 days: no wind, snow and fog no insolation; 1457

After those 10 days, normal weather for winter season; In normal operation algorithm, when battery is discharged (LVD react - disconnect battery), diesel back-up generator will feed the load and charge battery. When battery is fully charged, diesel generator stops, and load is fed from battery. When battery discharge, cycle repeats. In method proposed in this work, based on weather forecast, when diesel generator fully charge battery, load still will be power from diesel generator, not from battery. Reason for that is to reduce fuel consumption caused by efficiency of battery charging AC/DC convertor in every charging cycles. In last two days (of 10 days with no insolation) load will be fed from battery. After that, system goes back to normal operation. In black out period of 10 days, battery will have 4 to 6 cycles more if operating on traditional algorithm. For 5 cycles of battery charging, battery charging AC/DC convertor efficiency of 90%, 1200Ah battery, energy losses can be calculated approximately: E loss = 5 cycles x 80A x 48V x 9h x 0,1 = 17,289 kwh where 80A is average charging current during charging period, 48V is nominal battery voltage, 9h is charging period running on diesel generator and 0,1 present 10% of losses in AC/DC convertor. As explained in this chapter, control of battery charging current present key factor. Battery charger can be central, or separate, as it shown in Figure 2. Figure 2 shows intelligent control of hybrid power system described in this chapter, without central battery charger. CONCLUSION Realizing the fact that power energy is the key factor for sustainable development, a consensus has been reached among the scientists, energy engineers and economists as well as political communities all over the world that there is a need to supply energy for all sectors, including ICT without a detrimental impact on ecology. In this work, for telecommunication application has been suggested to use renewable energy in combination with traditional, to provide stable and reliable energy sources. After presenting of the components of described hybrid power system in this work and their sizing, authors present idea, basic principles and benefits of intelligent hybrid power system control through real life examples. Presented idea of intelligent hybrid power system control, for target has reducing of fuel consumption and CO 2 emission, decreasing of logistic and operational costs and cost effective component sizing as photovoltaic cell, wind generators, storage battery and diesel back-up generators. Future work will be focused on new working algorithm development. This algorithm will be base on real meteo data, which means number of variety and system adoption for short term and long term period. Generally, purpose of this work is to give some practical guideline for system design and component sizing and stimulate development of new idea in this area. REFERENCES [1] European Commision Repport, IP/09/1498, Brussels, 9. October 2009. [2] ETSI Technical Repport ETSI TR 102 532, The use of alternative energy solution in telecommunication installation. [3] E. Muljadi and J.T. Bialasiewicz, Hybrid power System with a Controlled Energy Storage, 29 th Annual Conference of the IEEE Industrial Electronich Society, Roanoke, Virginia, November 2003. [4] Seeling-Hochmuth, Optimisation of hybrid energy systems sizing and operation control, PhD thesis, Universiti of Kassel, October 1998. [5] E.F.E. Ribeiro, A.J. Marques Cardoso and C. Boccaletti Uninterruptible energy production in standalone power systems for telecommunication, International conference on renewable energies and power quality, Valencia, April 2009. [6] Reeve, DC Power system design for telecommunication, John Wiley and Sons, USA (2007). [7] Gumhalter, Power supply in telecommunications, Springer-Verlag, Berlin (1995). [8] D. B. Nelson, M.H. Nehrir and C. Wang, Unit sizing of standalone hybrid wind/pv/fuell cell power generator system, IEEE Power Engineering Society Meeting 2005, Vol. 3, pp 2116-2122, June 2005. Figure 2.Inteligent control of hybrid power system 1458