1. Introduction
The so-called French Reed Bed (FRB) is a particular constructed wetland (CW) solution which receives raw wastewater [
1,
2,
3]. FRB is a two stage system: the first stage involves a vertical subsurface flow (VF) bed receiving raw wastewater and filled with coarse gravel; the second stage is a VF bed filled with coarse sand. Primary treatments are not adopted, since the surface of the first stage VF acts as a filtration stage. Indeed, the solid materials from wastewater will create an organic top layer on the surface area and has to be removed after 10–15 years, i.e., when it is already stabilized and can be used as soil conditioner [
4]. FRBs are very efficient in the removal of suspended, dissolved organic matter, and pathogens and have a high nitrification capability, with a relevant contribution due to the first stage [
5]. Higher denitrification and total nitrogen removals are achievable with the adoption of a saturation bottom layer [
6]. The system does not present odor issues due to the fact that the sludge formed on the surface of the wetland is kept under constant aerobic conditions by the cyclic feeding scheme and the active rhizosphere growing in it. The main advantage of FRB is that it does not require the primary treatment system (septic tank or Imhoff tank), as requested by classical CWs. Consequently, FRB is an attractive solution to minimize the operational and maintenance (O&M) costs of wastewater treatment from small settlement.
FRB is being successfully applied in France, where up to now more than 4000 treatment plants are in operation (<4000 person equivalent—PE), with the oldest having almost 30 years of lifespan [
3,
7,
8]. Moreover, FRB has been also successfully applied for a big city in Moldova, which treats up to 20,000 PE [
9], and in tropical climates [
10,
11].
CWs are well-known to be able to reduce the O&M costs in comparison to classical technological solutions, such as activated-sludge systems [
12,
13,
14]. However, the particular sludge management of FRB systems allows O&M costs to be lowered even in comparison to classical CWs. Therefore, the aim of this paper is to highlight the following concept: FRB is a suitable solution to provide robust wastewater treatment as well as minimize O&M expenditures of WWTPs for small communities in comparison to activated sludge and classical CW systems. To this aim we discuss one of the first FRB wastewater treatment plants (WWTP) for domestic wastewater in Italy, i.e., the FRB WWTP of the Castelluccio di Norcia town (up to 1000 PE). The discussion is based on data collected from the Water Utility, which include: water quality monitoring; detailed estimations from the executive design; information from an interview to detail the real O&M costs faced by the FRB WWTP.
4. Discussion
The FRB WWTP of Castelluccio di Norcia was designed to face touristic fluctuation of produced wastewater. To this aim, the system is designed according to guidelines from French experience [
1,
3,
7,
8] (1.2 m
2 PE
−1 for the first FRB stage; 0.8 m
2 PE
−1 for the second stage) for the future peak of tourism, i.e., 1000 PE, with a slight undersize of the first stage to consider the seasonal fluctuation: 1 m
2 PE
−1 for the first stage FRB; 1 m
2 PE
−1 for the second stage VF. However, the system can be considered highly conservative for the touristic peaks faced during the monitoring period 2014–2016, assumable equal to 500 PE (2 m
2 PE
−1 for the first stage FRB; 2 m
2 PE
−1 for the second stage VF), and oversized for the off-season population of 200 PE (5 m
2 PE
−1 for the first stage FRB; 5 m
2 PE
−1 for the second stage VF). Although the size of the system should be considered in the interpretation of the water quality results, the average removal efficiencies of the FRB WWTP of Castelluccio di Norcia can be considered to be generally in line with the value reported from the ample dataset of French WWTPs (
Table 6). Even if the overall performance from Castelluccio reported in
Table 6 are from a three stage system (FRB + VF + FWS), the results are comparable with the data from the two stage systems analyzed by Paing et al. [
7] and Morvannou et al. [
8]. Higher nutrient removal resulted in comparison with French values (
Table 6). The higher TN removal can be attributed to denitrification in the third stage FWS, which is not considered (and usually not adopted) in French WWTPs. The FRB WWTP of Castelluccio di Norcia is a quite young WWTP in comparison to the systems analyzed by Paing et al. [
7] and Morvannou et al. [
8] (only 2 years old); therefore, the higher TP removal can be attributed to still not saturated adsorption sites for phosphorous. However, a part of the higher TP removal could be due to the effect of the third stage FWS.
Energy consumption is one of the expenditure items known for reducing the O&M costs of CWs in comparison to technological solutions. The FRB WWTP of Castelluccio di Norcia confirms this statement, with energy consumption in line with literature values reported for CWs. Assuming 200 PE for 5 months of off-season and 600 PE for peak touristic season to estimate the treated volume of wastewater (no measured data are available) and 150 L day
−1 PE
−1, the energy consumption results 0.15 kWh m
−3. This value is in line with the 0.1 kWh m
−3 for subsurface flow CW reported in literature, and one order of magnitude lower than energy needed from the most common technological solutions [
18]. The energy consumed by the WWTP of Castelluccio di Norcia is low, with low O&M cost. However, it must be noted that the majority of the energy costs are not due to consumed energy but to other costs, linked to energy network and fees. The other energy costs are probably so high, in comparison with the cost of energy, due to remote area in which the WWTP is sited and the low possibility, for the Energy Utility, to have income from the few activities connected to the electricity network. Therefore, the possibility to use renewable energy for WWTP functioning should be always considered in conditions similar to those of Castelluccio di Norcia, to reduce O&M costs not only in terms of cost of energy itself but principally for the linked cost to the service provider.
Another expenditure item in which classical CWs are known to be more advantageous in comparison to classical WWTP regards the sludge management. Essentially, activated sludge systems remove both particulate and dissolved organic load through sludge. Additional sludge is produced from activated sludge treatment plant if nitrification is required. Contrarily, CWs remove only the settable particulate organic matter as sludge within primary septic tanks. Indeed, the dissolved organic load in CWs is removed by biofilm attached to the porous media in subsurface systems, or by further settling of fine particle and biofilm attached to plant stems in FWS systems [
18], i.e., not contributing to sludge formation. Therefore, the amount of sludge to be disposed from classical CW systems is very low in comparison to that produced by classical activated sludge WWTPs, and consequently also the correlated costs. For instance, Masotti and Verlicchi [
13] reports for a small settlement of 300 PE in the Italian context a cost of 40 € PE
−1 year
−1 for sludge transport and disposal from classical activated sludge system, which is one order of magnitude higher in comparison to the value estimated by the same authors from the same settlement treated with classical CWs, i.e., 3.5 € PE
−1 year
−1. Regarding the issue of O&M cost reduction due to sludge management, the FRB solution represents a further improvement for CWs. Indeed, FRB system avoids septic tanks of classical CW schemes and accumulates the sludge on the top of the first FRB stage through the formation of a sludge deposit layer. The cracks produced by the movement of the plants with wind and the aeration pipes maintain the aerobic conditions within the deposit layer [
4], i.e., similarly to what happens within sludge drying reed beds [
3]. The oxic conditions are more favorable for the sludge mineralization than the anaerobic one developed in septic tanks, and the amount of sludge to be disposed at the end of a filling cycle of the first stage FRB freeboard is lower in comparison to that produced by classical CW schemes. Therefore, the FRB scheme adds two further advantages to classical CWs: (i) no need of yearly removal, transport, and disposal of sludge; (ii) lower volume of sludge to be removed, transported, and disposed during the overall lifecycle of the WWTP. These advantages contribute to a further decrease in O&M costs of FRB solution in comparison to classical CWs. The freeboard on the top of the first FRB stage at Castelluccio di Norcia was designed with a height of 0.4 m. The assumed growth rate of the deposit layer for FRB of Castelluccio di Norcia is 2 cm per year, slightly lower than the 2.5 cm per year suggested for FRB system [
4] to consider the fluctuation of the population due to touristic activities. Therefore, the freeboard is expected to be filled in 20 years. The transport and disposal of the accumulated sludge after 20 years is estimated to be equal to 8000 €, i.e., 400 € per year if distributed during the lifespan of the WWTP. Translated in terms of PE, the O&M sludge cost for the FRB WWTP of Castelluccio di Norcia results equal to 0.4–0.8 € PE
−1 year
−1 (1000 PE and 500 PE, respectively), i.e., one and two orders of magnitude lower than the costs for classical CWs and activated sludge reported by Masotti and Verlicchi [
13], respectively.
The FRB WWTP of Castelluccio di Norcia can be used to highlight the advantages of FRB scheme on the activated sludge system through the analysis of construction and O&M overall costs, which are reported in
Table 7. In terms of construction costs, it is proper to compare the cost of the FRB WWTP of Castelluccio di Norcia as dimensioned for 1000 PE (i.e., maximum treatment capacity of the WWTP);
Table 7 shows how the construction costs of the FRB WWTP (394 € per PE) were slightly higher but comparable with the construction costs of activated sludge systems in Italian context (263–360 € per PE). If the system would be realized strictly following the French scheme with only two stages (FRB + VF), the construction costs of FRB WWTP could be even lower. In this case, FWS was included due to restrictive water quality target requested to discharge on soil. If the FRB WWTP would be realized in area with less restrictive water quality limits (e.g., discharge in water body), the FWS could be avoided (about 30,000 €), leading to construction cost for the FRB scheme fully in line with higher range of activated sludge WWTP (364 € per PE). The FRB construction costs are in accordance with the value reported by Gikas and Tsihrintzis [
19] for a real WWTP in Greece, also designed with the FRB approach; the system discussed by Gikas and Tsihrintzis [
19] is designed for 600 PE, includes an additional third horizontal subsurface flow CW for denitrification and costs 477 € PE
−1. The FRB construction costs for the Castelluccio di Norcia WWTP are also in line with the value reported by Geenens and Thoeye [
20] for 1000 PE in Belgian context, both for CWs (430 € PE
−1) and activated sludge systems (380 € PE
−1). The ratio between construction costs of activated sludge systems and CWs is also in line with the analysis proposed by Batchelor and Loots [
12], which is based on a pilot CW study and is aimed for WWTP serving below 5000 PE in South Africa context; this study reports construction costs of CWs only 24% higher than those of activated sludge systems.
The O&M costs of the FRB WWTP of Castelluccio di Norcia must be considered for population faced during the monitoring period, i.e., assumable equal to 500 PE. Under this assumption, the O&M costs of the analyzed FRB WWTP results very low (11 € per PE) due to the advantages in terms of energy consumption and sludge management previously discussed. Comparing with classical WWTPs from Italian context, the O&M cost of the FRB WWTP of Castelluccio di Norcia results among 5 to 8 lower than classical activated sludge systems (
Table 7). It must be noted that the estimated O&M costs would not change significantly even if WWTP would face the maximum designed population of 1000 PE; among the considered 9 expenditure costs, the only one that is expected to change is the energy consumption for consumed kWh, while all the other activities and costs could be assumed to be done and spent in the same way for both 500 and 1000 PE (e.g., water quality samples, WWTP inspections, fees for energy network). Therefore, the O&M costs for 1000 PE reported in
Table 7 is estimated assuming all the expenditure item costs equal to those afforded for 500 PE, only doubling the energy costs per consumed kWh (additional 159 € year
−1–318 € year
−1 in total for 1000 PE). The result is reported in
Table 7 and shows an O&M cost per 1000 PE of 6 € PE
−1 year
−1, i.e., 8 to 13 lower than those of classical activated sludge systems. It must be noted that the O&M costs for the FRB WWTP of Castelluccio di Norcia are in line with the value reported by the Greece FRB real case study for 600 PE proposed by Gikas and Tsihrintzis [
19], who estimate an O&M cost equal to 12 € per PE. The calculated ratio between O&M of activated sludge system and FRB for Castelluccio di Norcia seems to confirm the capability of FRB system to minimize O&M in comparison to classical CWs. Batchelor and Loots [
12] reports O&M cost of CWs 4.6 lower than those of activated sludge solution (target 5000 PE). Masotti and Verlicchi [
13] estimated the O&M costs of activated sludge (prolonged aeration) 1.7 times those of CWs for a WWTP serving 300 PE. Therefore, the previous ranges report a saving of O&M costs due to the use of classical CWs instead of activated sludge all lower than the reduction of O&M costs calculated for the FRB WWTP of Castelluccio di Norcia; however, more comparison studies on both classical and FRB CWs with activated sludge system O&M costs are needed to confirm this trend.
The previously discussed O&M costs does not include any estimation of benefits due to additional ecosystem services provided by green instead of gray infrastructures [
21]. For instance, Ghermandi and Fichtman [
22] estimated a mean and median monetary flow due to recreational activities linked with FWS systems of 8397 and 530 € ha
−1 year
−1, respectively. Therefore, the O&M of FRB system could be even lower including also the natural capital revenues in the cost estimations.