**2 Literature Review**

**2.2 Ericsson background and their Power System Model**

Ericsson is a Swedish multinational networking and telecommunications company headquartered in Stockholm.

The company sells software, infrastructure and network services, such as 3G, 4G and 5G equipment to telecommunications providers. Besides, Ericsson is ranked highest in the 5G infrastructure market, making the Swedish firm a leader in the deployment of the new technology around the globe. Ericsson Enclosure & Power is a sub-division, based in the company headquarter in Stockholm, that develops a wide portfolio of cost-efficient power products for Telecom Sites, see Figure 12, such as enclosure, cabinets, inverter, genset, batteries and recently also solar panels, MPPT and wind turbines and wind inverters. Their main purpose is to provide reliable power to Telecom Sites and recently they have shown a growing interest for low-carbon alternatives, to reduce the telecommunications sector environmental footprint and to reduce the operating costs deriving from the use of fossil fuel based units like the genset in remote locations.

5g radio sites are expected to require three times as much energy as 4G but at the same time, software improvements are constantly improving the efficiency, decrease the power consumption of new antennas in term of kWh per gigabyte transmitted. 5G might eventually be consuming less than 4G by 2030. However, much more data will be required because traffic and network demand will steadily increase. Therefore, for the Jevon’s paradox, the overall energy requirements of 5G site will be higher, although it is difficult to understand to what extent.

15 Figure 12 - View of an Ericsson power enclosure at the bottome of a Telecom Site

Ericsson Enclosure & Power department is building a model tool on Excel to size and design hybrid energy systems to power their Radio sites such as antennas. The first phase of this thesis work consisted in contributing to add and improve some parts of it and this section will illustrate its structure and purpose. The electricity is generated from a combination of Solar PV and wind turbines in combination with a battery bank as storage system.

An optional secondary power is offered and the costumer can then decide among three different configuration:

stand-alone (or pure solar), diesel generator backup or grid connected system. As a result, the model produces a techno-economic evaluation of the chosen system, in terms of availability of renewable energy throughout the year, cost savings compared to an equivalent system connected only to the grid or to the gen-set and the payback period.

2.2.1 Solar and Wind production

The solar irradiance and wind data are collected from Soda [31] by just entering latitude and longitude of the location. The website will deliver meteorological data in terms of global horizontal irradiance (GHI) and windspeed for each hour of the time span chosen, usually one year. In the model, for every hour of the year, several parameters used to describe the angular position of the sun with respect to an oriented surface are calculated [26]. These are the declination angle (δ):

δ = 23.34 ∗ sin 360 ∗284 + 𝑛 365

Figure 13-declination angle [32]

16 with n the number of the day in the year, it describes the angular position of the sun respect to the plane of the equator at noon, it is the same in any location in the world, Figure 13. North positive; -23.45≤ δ ≥23.45. The hour angle (ω):

ω = 15 ∗ (h − 12)

where h is the solar time in hours, measures the deviation from south of the sun’s position, it is negative in the morning and positive in the afternoon and it changes 15 degrees every hour.

Figure 14 - hour angle [33]

The solar angle (SA), defined as the angular position of the sun compared to the horizon, is the complementary of the Zenith angle, as shown in Figure 15, and can be found in this way:

𝑐𝑜𝑠(𝜃 ) = sin(δ) ∗ sin(φ) + cos(δ) ∗ cos (ω)∗ cos(δ) 𝑆𝐴 = arcsin (cos(𝜃 ))

Figure 15-Zenith and solar angles [34]

The user can input the inclination of the PV surface as Tilt Angle (TA), based on which other factors can be computed. The Tilt Factor (TF), defined as:

𝑇𝐹 = cos|90 – SA – TA|

describe how much aligned the beam radiation and the tilted surface are, the more they are, the closer to 1 TF is.

The Delta Tilt Factor (DTF), is defined as the difference between the TF and the TF of a flat surface (for which TA=0). Since the irradiation data are given as global, it is necessary to distinguish between the components of direct and diffuse radiation. This is done through a linear approximation based on the fact that, when the sky is clear and the PV surface is perpendicular to the sun rays, the beam component is 90 % of the global irradiation and it decreases with 1/3 of the misalignment degrees

17 𝑃𝑜𝐷𝐼 =100 ∗ 0,9 − 1

3∗|90 – SA – TA|

100

Once differentiated the direct and diffuse component, the Tilt Effect (TE) is defined as:

𝑇𝐸 = 𝑃𝑜𝐷𝐼 ∗ (1 + 𝐷𝑇𝐹) + (1 − 𝑃𝑜𝐷𝐼) ∗ 90 − 𝑇𝐴 90 Finally, the total harvested energy from the PV (PVH) is calculated:

𝑃𝑉𝐻 = 𝐺𝐻𝐼 ∗ 𝑇𝐸 ∗ 𝑃𝐸 ∗ 𝑇𝑃𝐴 ∗ (1 − 𝑇𝐿)

Where PE is the panel efficiency, TPA the total area and TL total losses. One of the weaknesses of this method lies on the fact that the input data do not distinguish between direct and diffuse radiation but they consist only of global horizontal irradiation. Furthermore, in the PV harvest equation, when the TE is zero, the whole production goes to zero. The TE goes to zero when the solar angle is negative. However, during the sunset and sunrise hours, the sun is very close to the horizon and, since the calculations are made for an entire hour span, there could be situations in which the solar angle is negative but there is some radiation registered in the software, due especially to the diffuse component. This might lead to an error in the total harvested energy throughout one year, especially for high latitude locations, where the sun stays a relatively longer time near the horizon on a winter day.

In addition, the performance of PV depends on the spectral distribution of the incoming radiation and several studies have investigated the relation between that and the panels yield. Stark and Theristis [35] study the effect of atmospheric composition on the performance of a photovoltaic panel. Besides, it is known that the diffuse radiation has a different spectral distribution than the direct radiation [36], see Figure 16, the diffuse component highlighted in blue, is more concentrated among higher wavelengths, which are more effective in producing short circuit current [37]. Hence, considering just the overall incoming solar energy might distort the total output and distinguishing between direct and diffuse radiation from the beginning, would enable to evaluate their contribution to the PV yield differently, achieving more reliable results.

Figure 16 - Clear-sky direct and diffuse irradiances. Both spectra were normalised to have unit integrals. [38]

When a stream of air is in motion, it carries kinetic energy following the formula [39]:

𝑊𝑖𝑛𝑑 𝑃𝑜𝑤𝑒𝑟 =1

2∗ 𝜌 ∗ 𝐴 ∗ 𝑣

Where 𝜌 is the density of air in Kg/m3, A is the area of the air stream (or the area of the wind turbine rotor) in 𝑚 and v the wind speed in . The turbine blades convert the kinetic energy of the wind in mechanical energy and then the alternator in electrical power. The conversion efficiency is described by the Wind Turbine Power coefficient:

18 𝐶 =𝑒𝑙𝑒𝑐𝑡𝑟𝑖𝑐𝑎𝑙 𝑝𝑜𝑤𝑒𝑟 𝑜𝑢𝑡𝑝𝑢𝑡

𝑤𝑖𝑛𝑑 𝑝𝑜𝑤𝑒𝑟 𝑖𝑛𝑝𝑢𝑡

According to the Betz’s law [40], the maximum conversion factor from kinetic to mechanical energy that a wind turbine can achieve is equal to 0.593, when a 70% efficiency of the mechanical to electrical conversion is added, a value for Cp of around 0.41 is obtained. Typical values for Cp go from 0.30 to 0.45 [41], [42].

Figure 17-Example of power conversion in a wind turbine. The Betz limit says that no more than 59,3% of kinetic energy can be converted into mechanical power [41]

Then, by converting equation x with y the total electrical power harvested by the wind turbine is found:

𝐸𝑙𝑒𝑐𝑡𝑟𝑖𝑐𝑎𝑙 𝑃𝑜𝑤𝑒𝑟 𝑜𝑢𝑡𝑝𝑢𝑡 = 𝐶 ∗1

2∗ 𝜌 ∗ 𝐴 ∗ 𝑣

The tool offers different types of panels and wind turbines that the user can choose. There are four model of PV, among which a 455 𝑊 silicon monocrystalline and a 380 𝑊 , 355 𝑊 , 340 𝑊 silicon multicrystalline. Each of them differs for efficiency value and voltage and current parameters. With reference to the wind turbines, two models can be picked, a 5 𝑘𝑊 with a rotor diameter of 5 meters and a 9.8 𝑘𝑊 , whose rotor’s diameter is equal to 7.8 meters

2.2.2 Load

The load of the site can be added as an input, the model specifies a low load and a high load to give the option to distinguish between the load during the daytime and the one during the night time. The daytime is usually considered from 9am to 9pm. Ericsson sites requires a power input of about 120 W (simple LTE site) to 32 kW [43]. The main difference between Telecom Solar and other solar applications is that the latter are usually aimed at maximizing the energy harvest over time and that they can usually sell the excess to the grid. Telecom applications in pure solar mode require an over dimensioning of the panels, to harvest enough energy during the least sunny month and ensure the highest self-sufficiency level.

2.2.3 Battery Storage System

After modelling the power production and consumption, the energy storage system has to be designed. The tool offers several models of Lithium-ion and Lead-acid batteries, mainly differentiated by their nominal capacity, given in Ampere-hours and the charging efficiency, 95% for Lithium-ion and 85% for Lead-acid. Then, the user can input the amount needed and the total storage capacity can be simply calculated as:

𝑆𝑡𝑜𝑟𝑎𝑔𝑒 𝐶𝑎𝑝𝑎𝑐𝑖𝑡𝑦 [𝑊ℎ] = 𝑁𝑜𝑚𝑖𝑛𝑎𝑙 𝐶𝑎𝑝𝑎𝑐𝑖𝑦[𝐴ℎ] ∗ 𝑛 𝑜𝑓 𝑏𝑎𝑡𝑡𝑒𝑟𝑖𝑒𝑠 ∗ 𝑠𝑦𝑠𝑡𝑒𝑚 𝑣𝑜𝑙𝑡𝑎𝑔𝑒[𝑉]

Based on the electric power definition [44]:

𝑃 = 𝐼 ∗ 𝑉

19 where I is the current and V the voltage. Ericsson telecom site has 48 V as a default system voltage.

The utilization and charging scheme of the storage is controlled by a logic loop that includes a nested if. The code is used to calculate how much it is charged or discharged, to or from the battery bank, for every hour of the year.

However, instead of calculating the charge and discharge stage separately and combine them later, it directly computes the hourly storage level based on the power produced by renewables, power available from the secondary source, the load and the storage level of the previous hour.

2.2.4 Economics

In addition to the technical characteristics of the system, a financial analysis is performed by the model to evaluate the economic feasibility of investing in such a hybrid system. Firstly, a list of the capital investments of every component is made. These are:

Photovoltaic panels;

MPPT;

Batteries;

Wind Turbines;

the structure and the civil works costs are included as well. Besides, fixed operation and maintenance costs have to be addressed, together with any costs related to the replacement of some of the components, such as batteries, when they reach their limit of number of cycles. Then, two base cases are taken as references, one where the system is fully reliable on the grid and another in which it is supported only by the diesel generator. In each case, the yearly cost of generating energy is calculated, based on the retail electricity rate in USD/kWh the former and on the cost of the Diesel in USD/liters and the Diesel consumption of the system in liters/kWh for the latter. Next, the yearly generation from the renewable sources is subtracted from the total yearly energy demand and the cost of producing the remaining energy is calculated again. This cost will be lower than the previous one and their difference is considered as the savings of not using the grid or the diesel generator when the renewable energy is available. To analyse the profitability of such an investment that returns yearly savings, a discounted payback period method is used [45]. A discounted payback period is an economic tool that computes a yearly cash flow with the initial investment in the year zero, the operating costs and the savings produced every year, discounted by a certain interest rate, to keep into account the time value of the money. The investment will be considered economically feasible if the discounted cash flow will reach a positive value within the 25 years of operation considered.