Keywords

1 Introduction

Air pollution is known to be a serious problem mainly for its effects on human health. Recently, however, the air pollution effects in other fields such as the photovoltaic energy generation have been investigated and, notably, the relation between the reduction of the solar insolation reaching the PV systems and fine particular matter (PM2.5) concentration in the air. Air pollution typically includes PM2.5 which suspended in the atmosphere can reduce the solar radiation intensity reaching the ground. At this regard, Peters et al. [1], correlating measured particulate concentrations and solar insolation in Delhi over a long period of 19 months, estimated that total sunlight reaching the ground during one year was reduced by more than one ninth, due to air pollution.

Wang et al. [3], testing a distributed photovoltaic system on a building roof in Shanghai, showed that the higher the PM2.5 concentration, the lower the power generation capacity of the PV module was. Then, the solar radiation received on the surface of the PV module resulted exponentially related to the atmospheric PM2.5 concentration.

It is known that the cities in the Eastern Mediterranean and Southeast Asia are often affected by major haze events reaching PM2.5 concentrations up to 375 μg m−3 [1, 2]. However, most of cities around the world suffers the air pollution although at several seriousness levels.

In this study, we intend at addressing the impact of PM2.5 particles on solar insolation levels in the city of Naples (South Italy), whose air pollution levels exceed often the threshold values.

By 2030, EU countries will have to increase the use of renewables, including the solar PV systems, for realizing the 32% target of renewable energy production, according to the EU’s re-cast renewable energy directive [4]. In this regard, EU orientation is to address the use of suitable buildings’ surfaces rooftops and facades for distributed solar PV systems deployments.

Recently, Bodis et al. [5] have estimated that the EU cities’ building rooftops could potentially produce solar PV electricity, annually covering the 25% of the current electricity. In particular, Italy could potentially cover more than 30% of its electricity consumption by developing rooftop PV systems at its most advantageous rooftops. In view of exploitation of this potential, the estimation of the reduction of solar radiation due to PM2.5 particles in the air could be fundamental, making the difference between a solar PV installation meets the expected output and one that fails. In this study, at this scope, an empirical relation between PM2.5 concentration and reduction in insolation has been derived by analyzing one-hour PM2.5 and insolation measurements from a monitored location in Naples, Italy.

2 Correlating PM2.5 Concentration and Reduction in Solar Radiation

The collected data used in our analysis were recorded over 19 months between January 2018 and July 2019. Insolation data were measured by a pyranometer with a frequency of one measurement every hour. This instrument consists in a thermopile sensor coated with an opaque black paint providing a flat spectral response for the full wave length range. It measures the global solar radiation on a plane/level surface as sum of direct solar radiation and diffuse sky one.

Fine particulate data were recorded by an air quality monitor as SWAM 5A DUAL CHANNEL with one-hour frequency.

Both instruments are located at the air quality gauge station, located in Naples, managed by ARPAC (Italian Agency of Environmental Protection). This station is identified by the EU code IT1493A.

The approach, used for relating PM2.5 concentration and insolation data, consists in three main steps: (1) normalizing insolation data, (2) clustering PM2.5 concentration data, (3) deriving the correlation curve.

At first, PM2.5 concentration levels were classified according the color coding of Air Quality Index (AQI). We observed that most of recorded concentration levels fell in the first AQI class ranging among 0–50 μg m−3. For making the analysis consistent, we split the first class in sub-classes as showed in Table 1, transferring more properly the recorded data in AQI levels.

Table 1 PM2.5 concentration ranges and corresponding AQI color code
Full size table

Hence, insolation data were sorted in bins corresponding to the defined different PM2.5 concentration levels. Then, humidity and clear sky filters were used for identified data representative of clear sky conditions.

It is to be noted that, over the course of a year, the insolation varies via the zenith angle and the eccentricity of the Earth’s orbit around the sun as well as due to seasonal variations in atmospheric conditions. Since insolation data covered a longer period than one year, these variations were to be considered in the correlation analysis. At this scope, insolation data were analyzed for each month separately and normalized by taking the 90 percentile value of all insolation measurements at noon for all months.

Figure 1 shows the results of the normalization for the most representative months of June and November.

Fig. 1
figure 1

Graphs of insolation data. On the left, the raw data and on the right the data normalized, corresponding at June and November months

Full size image

For identifying the conditions representative of a clear sky, the 80 percentile filter of the datasets for each hour and pollution level was calculated.

This data was used for deriving the correlation curve between PM2.5 concentration levels and reduction of insolation, considering the relative reduction for each hour. For generating this functional relation, only datasets containing more than seven data points were considered.

The data analyzed were fitted by an exponential decay curve with a value R2 of 55%.

Exponential decay was expected according to Lambert–Beer’s law. The fitted exponential decay is shown as a black line in Fig. 2.

Fig. 2
figure 2

Exponential variation of solar radiation with PM2.5 concentration

Full size image

The fitted curve is described by the following equation:

$$\frac{{I(PM_{2.5} )}}{{I_{o} }} = \exp \left( {\frac{{ - PM_{2.5} }}{250}} \right)$$
(1)

where I0 is the isolation at 0 μg m−3 and I is the insolation affected by PM2.5 concentration.

As simplification, this result is obtained assuming consistent, over the entire period considered, the composition and size distribution of air pollution as well as the optical behavior of the aerosol.

3 Estimating PM2.5 Related Reduction in Solar Insolation

Equation (1) was used to estimate over one year how much light is lost due to PM2.5 related air pollution in Naples.

Figure 3 shows the measured insolation and the insolation estimated by means of Eq. (1), considering PM2.5 concentration of 0 μg m−3 (I0), over one year from May 2018 to May 2019. Integrating this insolation data, we calculated the annual insolation and the projected insolation at 0 μg m−3 PM2.5, as shown in Table 2.

Fig. 3
figure 3

Measured insolation and estimated insolation without air pollution due to PM2.5 over one year

Full size image
Table 2 PM2.5 effects on solar insolation in Naples
Full size table

So, we found that the amount of insolation for Naples was reduced by 5% or 66.20 kWh/m−2 of the annual solar energy reaching the ground.

From this result, we can derive that the air fine particulate pollution at concentration levels of 0–100 μg m−3 can affect the solar energy potential of viable rooftops for PV installations. Moreover, using this empirical relation, the insolation loss due to air quality can be directly evaluated, making more reliable the solar PV potential estimations [6].

4 Conclusions

Atmospheric particulate pollutant can affect the transition of light through the lower atmosphere, reducing the solar radiation reaching a PV panel. So, the variation of solar radiation intensity with PM2.5 concentration, mainly in the cities, is investigated in order to make more reliable PV power estimations.

In this work, we have analyzed the correlation between PM2.5 concentration and the loss in isolation, using long term field data from a monitored location in Naples. As result, we obtained that the amount of solar radiation is exponentially correlated to the PM2.5 concentration.

Using the formulated functional relation, we calculated the amount of insolation Naples would have received without the air particulate pollution within one-year period. In particular, we estimated that the insolation due to PM2.5 pollution was reduced around 5%, from 1368 to 1302 kW m−2.

The derived empirical relation represents itself a result. It enables a way to predict the influence of air fine particulate pollution on the solar energy potential in different PV suitable sites.

In conclusion, this study provides the theoretical basis to design solar PV systems to be mounted on a building rooftop as well as in other suitable sites, taking also into account the local air pollution condition.