Integrated Framework to Assess Advanced Phosphorus Recycling as a Sustainable Alternative to Sewage Sludge in Agricultural Soils

Abstract

Advanced phosphorus (P) recycling from wastewater is critical for improving nutrient circularity and reducing soil pollution associated with the direct application of sewage sludge in agriculture. However, few studies evaluate the long-term environmental and economic trade-offs between recycled P products and raw sewage sludge application. This study compares struvite, vivianite, and dicalcium phosphate (CaP) as P alternatives to sludge to mitigate heavy metal accumulation in Spanish agricultural soils. Using data from 27,835 plots, heavy metal accumulation was simulated over 50- and 100-year fertilisation scenarios. The results indicate that continuous sludge application leads to widespread exceedances of zinc, copper, and cadmium, especially in alkaline soils, whereas substitution with recycled products can substantially reduce these risks. Vivianite balances P recycling and costs, CaP offers the best environmental performance but with higher investment, and struvite suits smaller regions prioritising environmental safety. Economic analysis favours advanced recycling over sludge, especially considering externalities such as soil remediation costs. Despite limitations, our findings emphasise the importance of integrating environmental externalities into economic assessments and the value of advanced P recycling for sustainable soil management.

1. Introduction

Phosphorus (P) is an essential element for food production and crucial for multiple industrial applications. In the current context, phosphate rock is the only reliable P source, 85% of which is used to produce mineral P fertilisers for agriculture [1]. However, phosphate rock is finite and unevenly distributed across the globe, thereby increasing the risk of food supply disruptions in highly dependent regions [2]. To address this issue, recycling P from waste has been highlighted as an alternative to mineral P [3], where recycling refers to the reintroduction of P contained in waste back into an earlier stage of the cycle [1]. Among several P-containing wastes, sewage sludge, a by-product of urban wastewater treatment plants (WWTPs), has emerged as a promising P source for its relatively high P content (15–25 g·P kg⁻¹ of dry matter), representing over 80% of the P inflow of WWTPs in case of enhanced P removal [4].

Although sewage sludge reuse in agriculture presents a possibility for P recycling, it comes with significant challenges. Besides nutrients, sewage sludge also contains heavy metals (e.g., cadmium, lead, and mercury), organic compounds (e.g., pesticides, per- and polyfluoroalkyl substances (PFAS), pharmaceuticals), pathogens, microplastics, among others, all of which pose risks to human and environmental health [5,6]. Moreover, the long-term fate of metals from sludge remains largely uncertain because gradual inputs can accumulate in soils over decades, with persistence and bioavailability strongly modulated by soil pH, organic matter, and redox conditions (refs). For this reason, the direct application of raw sewage sludge in agriculture, while extensive in some countries (e.g., Spain; [7]), has been progressively restricted in many others (e.g., Germany, Austria and Switzerland; [8]).

Advanced technologies have been developed to recover P from sewage sludge while reducing potential contaminants, resulting in environmentally safer fertilisers, more commonly calcium phosphates (e.g., dicalcium phosphate (CaP)), struvite (magnesium ammonium phosphate), or vivianite (iron phosphate) [9,10,11]. These recycled P sources can be used as fertilisers in agriculture, with agronomic performance depending on the product characteristics, environment and management practices [12,13]. Recycled products have been shown to exhibit similar efficiency to mineral P fertilisers and can lower the risk of introducing pollutants (i.e., heavy metals) into agricultural soil compared to the direct application of sewage sludge [10]. However, these technologies require substantial capital and operational costs and, according to previous economic analyses, remain financially less viable than direct sewage sludge application or mineral P fertilisation [9].

Existing economic assessments of P recycling focus solely on direct costs and benefits of implementing a new technology, such as disposal or maintenance cost reductions and revenue from product commercialisation [9,14]. Although some studies have successfully included externalities of less sustainable processes to offset the costs of implementing innovative technologies, these lack integration of technical environmental impact evaluation to compare advanced P recycling technologies and support decision-making [15]. A truly integrative cost–benefit analysis (CBA) should also consider the environmental externalities of direct sludge application in agriculture, particularly the risk of heavy metal accumulation in soils, which could be mitigated through the adoption of advanced, safer recycled P products [8].

To address this gap, we propose a CBA framework that integrates the estimation of the environmental impact of direct sewage sludge application in agricultural soils with a CBA of implementing three different P recycling technologies, accounting for externalities such as heavy metal accumulation in soil. Factors such as sludge application rates and soil characteristics (e.g., pH, P status and heavy metal concentrations) vary among regions [16,17]. Therefore, our methodology serves as a model for comparing the environmental and economic performance of P recycling technologies for local agricultural and wastewater treatment infrastructure. By taking Spanish regions as an example, we highlight the importance of incorporating environmental externalities into policy and decision-making considering different regional conditions.

2. Materials and Methods

The framework is delimited by the pathway from wastewater treatment to agricultural soil application (Figure 1). Sewage sludge generated at wastewater treatment plants can either be directly applied to soil or processed into alternative P fertilisers, with struvite, vivianite and CaP holding particular importance [12]. We estimated the heavy metal accumulation and economic performance of the three aforementioned recycled P fertilisers and compared them with sludge, which represents the standard practice in Spain.

Figure 1. Schematic representation of system boundaries for the use of P from wastewater sewage sludge: current sewage sludge P (baseline scenario) and alternative recycled P (struvite, vivianite, and CaP).

The framework accounts for direct costs, including energy use, chemical agents, operator labour, transportation, and capital expenditure (CAPEX) associated with the production of the recycled fertilisers. Direct benefits were represented by P savings through avoided mineral fertiliser use, and revenue from potential P surplus sales. Environmental externalities were incorporated in the model by including the cost of soil remediation in contaminated areas and the opportunity cost arising from foregone crop production when sludge application is restricted. Collectively, these elements form the analytical basis for evaluating the integrated environmental and economic performance of P recovery strategies at the regional scale in Spain to promote sustainable soil management.

2.1. Sewage Sludge Application and Heavy Metal Content in Soil

In a first step we estimated the potential heavy metal accumulation resulting from scenarios involving the application of four P fertilisers in Spanish agriculture: sewage sludge, struvite, vivianite, and CaP, over 50- and 100-year periods. Simulation periods of 50 and 100 years were selected to represent medium- and long-term management horizons commonly used in soil contamination and nutrient balance studies [18,19]. The 50-year horizon approximates the typical timescale of agricultural soil use under consistent management, whereas the 100-year period enables assessment of potential legacy effects from slow-accumulating heavy metals.

This analysis is based on plot-specific soil characteristics, the composition of sewage sludge and P recovery products, as well as an accumulation model and applicable regulatory thresholds.

2.1.1. Sewage Sludge Application in Spanish Agriculture

The use of sewage sludge in agricultural soil is regulated by Spanish Law 1310/1990, which includes guidelines for sewage sludge application and reporting. This information was compiled by the Spanish Ministry of Ecological Transition and Demography in a database from 2018, containing the physicochemical (e.g., heavy metal content, pH, P content) and microbiological characteristics of the sludge used in over 28,000 agricultural plots across all Spanish regions [20]. The database also provides detailed information about the plots such as crop type, coordinates, application area, amount of sludge applied, and WWTP details (code, capacity and location) that supplied the sludge (Table 1). However, only the Spanish regions Andalusia, Castile-La Mancha, Castile-Leon, Catalonia, Extremadura, Galicia, Madrid, Murcia, Rioja and Valencian Community were included in our study, as the nine remaining regions (Asturias, Cantabria, Pais Vasco, Navarra, Aragon, Balearic Islands, Ceuta, Melilla and Canary Islands) did not report the use of sewage sludge in agricultural soil.

Table 1. Data on sewage sludge application across Spanish autonomous communities, detailing number of plots, area in ha (mean of 5.9 ha per plot), mass in t (mean of 5.5 t of sludge per plot), and mean heavy metal concentrations (Cd, Cu, Ni, Pb, Zn, Hg, Cr, P) in mg·kg⁻¹ sludge.

Community	Plots	Area	Mass	Cd	Cu	Ni	Pb	Zn	Hg	Cr	P
Andalusia	11	877.0	253.5	1.0	268.0	19.8	25.4	554.5	0.3	20.2	8897.2
Castile-La Mancha	10,777	105,903.6	57,773.5	1.5	260.6	35.7	41.9	779.0	5.3	73.6	20,823.6
Castile-Leon	133	1885.5	12,608.1	1.8	364.4	52.7	100.0	1367.5	0.7	123.0	13,282.6
Catalonia	13,804	9962.0	35,706.7	1.3	470.9	61.9	59.2	1275.0	1.1	110.9	27,631.9
Extremadura	409	20,634.3	24,939.4	1.2	206.6	24.0	43.3	511.2	0.5	38.2	10,736.7
Galicia	3	69.7	573.7	0.3	13.0	10.0	1.3	63.3	0.0	48.7	207.0
Madrid	85	941.3	1419.0	2.1	361.2	30.8	44.7	541.1	0.0	101.1	24,505.7
Murcia	1452	19,966.1	12,595.7	1.6	239.4	15.3	33.8	660.5	0.3	35.9	19,213.6
Rioja	1154	3479.8	6773.7	1.9	239.9	44.6	45.2	1111.6	0.4	92.5	15,847.6
Valencian Community	7	17.0	45.0	2.0	401.0	21.0	44.0	1616.0	1.0	33.0	10,469.9
National average				1.4	362.9	47.8	50.5	1030.8	2.6	90.7	23,727.2

2.1.2. Soil Characteristics of Plots

Soil properties of each plot were obtained using the Zonal Statistics Tool of QGIS to analyse the raster files from the European Soil Data Centre database and extract the values of cadmium (Cd), copper (Cu), nickel (Ni), lead (Pb), zinc (Zn), mercury (Hg), chromium (Cr) and pH at the exact coordinates [21,22]. In the case of exact coordinate points, the Zonal Statistics tool uses each plot coordinate as a ‘zone’ and samples the coincident cell value from the raster files (Cd, Cu, Ni, Pb, Zn, Hg, Cr and pH), extracting the exact value. Some plots (n = 292) presented a surface area of zero, likely owing to rounding or loss of decimal precision during file-type conversion, thus these records were excluded from the analysis. Similarly, plots with a reported sludge application mass of zero (n = 604) were removed, resulting in a final dataset comprising 27,835 plots. In some plots, the points were rendered outside the raster maps (28 points for Zn and 216 for pH), and these values were replaced with the overall average (Figure 2).

Figure 2. Sample group of agricultural plots (black triangles) where sewage sludge is applied and were rendered outside the Zn raster layer (n = 28).

2.1.3. Scenarios and System Boundaries

Heavy metal pollution resulting from sludge-based P fertilisation originates at the WWTP, and its environmental impact may extend to water bodies [23,24,25]. However, the transfer of heavy metals from soil to water bodies primarily occurs through leaching and runoff, which are highly variable flows that depend on location-specific factors, such as soil characteristics, slope, precipitation, and proximity to ground or surface water [26]. Consequently, our analysis focuses on the long-term heavy metal accumulation in soil and its economic implications for regional Spanish agriculture.

Before sewage sludge can be used in agriculture (Table 2), it is in most cases treated to reduce pathogens and odours [27,28]. This includes stabilisation to reduce organic matter and odours, usually through anaerobic or aerobic digestion [29]. The sludge is then dewatered to lower its moisture content, making it easier to handle and transport for application to agricultural soil [29].

Table 2. Comparative summary of sewage sludge and alternative phosphorus recycling technologies [30,31,32].

Phosphorus Fertilisers	Source Stream	Recovery Efficiency	Advantages	Disadvantages
Sludge (baseline scenario)	-	-	No investment required	High heavy metal content
Struvite	Sludge liquid phase	>95 % (~30% P inflow)	High agronomic value fertiliser	pH sensitive, high operational costs
Vivianite	Sludge	>80 % of vivianite-bound P (~60–65 % P inflow)	Low operational costs	High chemical consumption (additional iron salts)
CaP	Sludge ashes	>95 % (~80% P inflow)	Removes additional pollutants (e.g., microplastics)	High capital investment

In contrast to direct application, P can be recovered from sludge using different chemical and physical processes. Struvite, vivianite, and CaP can be produced depending on the treatment method, chemical dosing, and infrastructure in place at WWTPs (Table 2) [32]. First, after dewatering the sludge, the liquid phase (or press water) is recirculated back to the treatment step and contains up to 20–30% of P inflow to WWTP. This press water can be precipitated as struvite by dosing a magnesium salt (typically magnesium chloride) into a crystallisation reactor [33,34]. Alternatively, P in sludge binds with iron to form vivianite under anaerobic conditions when sufficient iron is present in the sludge, by dosing iron salts (typically iron chloride). Through magnetic separation, 60–65% of the total P in the WWTP inflow can be recovered [11,35]. Yet another possibility is the incineration of the stabilised and dewatered sludge [36], producing sludge ashes. These ashes contain up to 80% of the P inflow to the WWTP and can be chemically treated (acid leaching) to recover P with over 95% efficiency [37].

The varying chemical characteristics of the four P fertilisers studied provide advantages or disadvantages depending on soil conditions. In the context of heavy metal accumulation, their heavy metal content is a crucial factor (Table 3). In our framework, each scenario refers to the long-term application of a specific fertiliser product, where the baseline scenario corresponds to the current practice of sewage sludge use, and the alternatives correspond to struvite, vivianite, and CaP.

Table 3. Phosphorus and heavy metal content of studied fertilisers [11,18,38].

	P	Cd	Cu	Ni	Pb	Zn	Hg	Cr
	g·kg−1				mg·kg−1
Sludge—national average (baseline scenario)	23.7	1.4	363	47.8	50.5	1031	2.6	90.7
Struvite	169	0.4	19	5	6	260	0.1	14
Vivianite	104	0.4	130	42	15	380	0.2	120
CaP	169	0.1	5	2.5	3.6	15	0.1	1.7

2.1.4. Heavy Metal Accumulation and Soil Limits

In Spain, the permissible concentrations of heavy metals in agricultural soils are regulated by the Spanish Soil Nutrition Decree 1051/2022, which distinguishes between acidic (pH < 7) and alkaline (pH ≥ 7) soils (Table 4; [27]). Accordingly, the heavy metal limits varied among plots depending on soil pH.

Table 4. Heavy metal concentration limits for agricultural soils with different pH according to Spanish legislation [27].

	Limit Values (mg·kg−1 Soil)
	pH < 7	pH ≥ 7
Cd	20	40
Cu	1000	1750
Ni	300	400
Pb	750	1200
Zn	2500	4000
Hg	16	25
Cr	1000	1500

The heavy metal accumulation was estimated as the annual change in soil concentration for each plot ($i$) from yearly ($t$) heavy metal input of the respective P source applied (Equation (1)) as follows:

$${c_{soil}}_{i,t+1}={c_{soil}}_{i,t}+{\displaystyle \frac{{m_{input}}_{i,t}}{M_{soil}}}$$

(1)

where the mass input of a heavy metal (${m_{input}}_{i,t}$ in g·ha⁻¹) spread across a certain soil mass ($M_{soil}$) was added to the actual concentration of that heavy metal for year $t$ (${c_{soil}}_{i,t}$) to obtain the accumulated concentration for the following year $t+1$ (${c_{soil}}_{i,t+1}$). We assumed accumulation in the top 0.25 m layer of soil, with a bulk density of 1.2 t soil·m⁻³ resulting in a soil mass of 3000 t soil·ha⁻¹ [39].

Soil leaching and plant uptake were not included in the accumulation model, as their contribution to element output has been shown to significantly decline with increasing soil pH and decreasing precipitation [18]. This is consistent with national conditions in Spain, except in some of its regions, namely Galicia, Asturias, Cantabria and Pais Vasco [21,40]. For this reason, and to ensure consistent conditions across all fertilisation scenarios, we applied a mass balance that focuses on differences in heavy metal inputs between sludge, struvite, vivianite, and CaP.

2.2. Cost–Benefit Analysis

The cost–benefit analysis evaluates the economic viability of substituting sewage sludge with advanced P recycling technologies by incorporating direct costs and benefits, environmental externalities, and CAPEX. It also includes the calculation of the net present value (NPV) and the benefit–cost (B/C) ratio to support policy-relevant comparisons between treatments.

2.2.1. Capital Cost of Phosphorus Recycling Technologies

In our cost–benefit framework, the CAPEX for P recycling was annualised over 25-year cycles with full reinvestment at the end of each cycle. For the 50- and 100-year horizons considered, this assumption implies one and three full reinvestments, respectively. In the baseline scenario of sewage sludge application, existing wastewater treatment infrastructure was assumed to be already available during the first 25 years, so CAPEX was set to zero for this initial period. After year 25, reinvestments of the baseline infrastructure were also included, consistent with the alternative scenarios.

For the recycling technologies, there are references regarding the CAPEX required for a specific WWTP size (see Table S2) [9,41,42,43]. Through the WWTP code, we obtained the capacity of all the WWTPs that provided sludge for agricultural use in the registered plots. However, most of the WWTP sizes were different from the reference, and thus the CAPEX was adjusted to the actual scale of the WWTP through a scale–CAPEX relationship (Equation (2); [44]):

$$CAPEX={{CAPEX}_r·\left({\displaystyle \frac{S_r}{S}}\right)}^{0.6}$$

(2)

where ${CAPEX}_r$ and $S_r$ correspond to the CAPEX required for a reference WWTP size, while $CAPEX$ and $S$ represent the CAPEX required for the actual WWTP size that provided the sludge for agricultural use.

In the case of the alternative recycling technologies, total CAPEX included both the baseline wastewater treatment infrastructure and the additional technology required for P recovery. This ensures comparability across scenarios by assuming that all WWTPs met the same baseline requirements.

Most of the WWTPs supplying sludge for agricultural use applied only part of their total sludge production. In contrast, implementing P recycling technologies requires investment at the scale of the full WWTP, since recovery units treat the entire sludge stream rather than the marginal share applied to agricultural plots. As a result, the recovered P generally exceeded the regional agricultural demand, creating a surplus that could be commercialised. Therefore, for the recycled P scenarios, we included the sales revenue of P surplus (if available) within the benefits described in Section 2.2.2.

2.2.2. Direct Costs and Benefits

The direct benefits ($B_t$) were composed of the savings from mineral P replacement and, in the case of recycled P products, sales revenue from P surplus (Equation (3)). To account for these benefits, we assumed a price of 2000 EUR t⁻¹·P, based on the commercial price of triple superphosphate, a commonly used mineral P fertiliser [9].

$$B_t=\left(p_P·m_{P_{used}}\right)+\left(p_P·m_{P_{surplus}}\right)$$

(3)

where $m_{P_{used}}$ is the mass of P applied to soils, $p_P$ the price of P, and $m_{P_{surplus}}$ the surplus recycled P available for sale.

In contrast, direct costs ($C_t$) included the costs of treatment ($C_{treatment}$) (e.g., energy, chemicals, waste disposal), transportation ($C_{transport}$), and annualised CAPEX ($CAPEX$) [9,42] (Table S2):

$$C_t=\left(C_{treatment}·m_{P_{used}}\right)+\left(C_{transport}·d·m_{P_{used}}\right)+CAPEX$$

(4)

where $m_{P_{used}}$ denotes the mass of P applied to soils.

2.2.3. Externalities

Externalities in the CBA were represented by the costs of soil pollution with heavy metals. Specifically, we accounted for soil remediation costs (${C_{rem}}_i$) when legal thresholds were exceeded, and opportunity costs (${C_{op}}_i$) from lost crop production in contaminated plots, that is, the cost of not being able to produce agricultural products due to the exceedance of heavy metal legal limits [45]. We incorporated remediation costs at the year ($t)$ when heavy metal limits were first exceeded in each plot, while the opportunity cost was applied annually starting in the same year as the exceedance.

Remediation costs vary (10–1000 EUR t⁻¹·soil) depending on the technique [43]. We considered the in situ phytoremediation technique, a commonly used technique for heavy metal contamination remediation [46], costing around 100 EUR t⁻¹·soil [43]. The crop prices were obtained from the Spanish national accounts [47,48] (Table S1).

$${C_{rem}}_i=C_{rem}·A_i$$

(5)

$${C_{op}}_i=p_{crop}·A_i$$

(6)

where $p_{crop}$ is the price of the specific crop grown at each plot per hectare and $A_i$ is the plot area.

2.2.4. Economic Performance and Sensitivity Analysis

To evaluate the economic case, we used two commonly applied indicators of efficiency for each phosphorus fertilisation scenario in region j: Net Present Value (NPV) in million euros (EUR) and the benefit–cost (B/C) ratio.

$$NP{V}_j={\sum }_{t=0}^{T-1}{\displaystyle \frac{B_{j,t}-\left(C_{j,t}+CAPE{X}_{j,t}+C_{op,j,t}+C_{rem,j,t}\right)}{{\left(1+r\right)}^t}}$$

(7)

$$B/C_j={\displaystyle \frac{{\sum }_{t=0}^{T-1}\frac{B_{j,t}}{{\left(1+r\right)}^t}}{{\sum }_{t=0}^{T-1}\frac{C_{j,t}+CAPE{X}_{j,t}+C_{op,j,t}+C_{rem,j,t}}{{\left(1+r\right)}^t}}}$$

(8)

where $B_{j,t}$ and $C_{j,t}$ are the direct benefits and costs in year $t$, $CAPE{X}_{j,t}$ is the annualised capital cost of infrastructure, $C_{op,j,t}$ represents the opportunity cost, which was zero until the year of exceedance and positive thereafter, and $C_{rem,j,t}$ denotes the remediation cost, which was zero in all years except the specific year of exceedance, when it was applied once.

The NPV quantifies the difference between the present value of benefits and the present value of all related costs over time. Therefore, a positive NPV suggests that the advantages (e.g., avoided remediation costs) outweigh the investments associated with the use of a given fertiliser [49]. The B/C ratio reflects the relative efficiency of the investment, where values greater than one indicate that benefits exceed costs, and values below one suggest the opposite [49]. In our context, interpreting these indicators enables us to assess P recycling not as a purely commercial return, but as an environmental service, where slightly negative NPVs can still be worthwhile if they deliver lasting benefits to soil health, reduced pollution, and agricultural sustainability.

For projects with potential long-term impacts on society and the environment, the European Commission recommends its Member States and Cohesion countries to consider a 3–5% discount rate [50]. To provide a conservative scenario, we considered a 5% discount rate. In addition, to assess the robustness of the economic evaluation, we conducted a sensitivity analysis by varying the discount rate ($r$) from 3% to 7% for the NPV and B/C ratio for each P fertilisation scenario.

In terms of implementation, the sludge scenario assumed no capital expenditure during the first 25 years, since existing wastewater treatment infrastructure was already in place. From year 25 onwards, the baseline annuity was applied. For struvite, vivianite, and CaP, technology-specific annuities were included from the beginning, and both technology and baseline annuities were applied from year 25 onwards. Opportunity costs were accounted for annually starting from year 25, while remediation costs were charged only once at the end of the planning horizon.

2.3. Limitations of the Methodology

There are limitations inherent to the methodology that could introduce inaccuracies in estimates of heavy metal accumulation and economic performance. Uncertainty in long-term soil heavy metal behaviour was not quantified explicitly. In particular, the exclusion of leaching and plant uptake introduces uncertainty regarding the exact timing of threshold exceedance. However, such processes depend on location-specific unpredictable factors such as rainfall, which can vary substantially over time. Moreover, we simplified the estimation of non-financial costs of heavy metal accumulation, excluding the costs of health risks from contaminated crops, as well as further environmental impacts such as water pollution [10]. Our framework is limited to the economic effects of heavy metals in agricultural soils resulting from sludge input, while further pollutants present in sludge such as PFAS, pharmaceuticals, pathogens and microplastics could pose further environmental and health risks [51,52]. Conversely, the framework does not account for potential agronomic benefits of using sludge in agriculture, such as increased soil organic carbon and inputs of additional nutrients [6]. While the extent of these inaccuracies cannot be precisely quantified, they likely affect both costs and benefits, and their omission should be considered when interpreting the results, especially in borderline cases.

3. Results and Discussion

3.1. Heavy Metal Accumulation in Agricultural Soils

Most plots exhibited alkaline or neutral soils, particularly in Catalonia, Castile-Leon, Murcia and Rioja (13,316, 10,344, 1445 and 1151 plots, respectively) (Figure S1). Acidic soils were mostly found in plots from Catalonia (n = 488), Castile-La Mancha (n = 433) and Extremadura (n = 315). In contrast, Andalusia, Rioja, Valencian Community and Murcia showed minimal or no use of sewage sludge on acidic soils, while in Galicia all plots had acidic soils. This distribution reflects regional variability in soil properties and agricultural practices, both of which influence sludge application and heavy metal accumulation. Increased soil pH generally reduces dissolved Cd, Cu, Ni, Pb, and Zn via adsorption and carbonate/hydroxide formation, lowering mobility and plant availability, and thus, favouring in-soil stock build-up under repeated inputs. This immobilisation leads to lower leaching risks in alkaline soils, while at the same time, contributes to the accumulation of heavy metals. Conversely, in acidic soils, higher heavy metal mobility leads to more losses and uptake, leading to lower accumulation in the long term [53].

3.1.1. Soil After Direct Sewage Sludge Application

Our results highlight the complex interplay between long-term sludge application, heavy metal accumulation, and soil pH, emphasising the need for region-specific management strategies (Figure 3). After 50 years of sludge input, most plots still remained within regulatory limits, but approximately 9% exceeded at least one heavy metal threshold (n = 2447), with a significant portion (n = 251) surpassing the limits for multiple metals. Soil pH strongly influences metal retention and mobility, with acidic soils typically promoting higher metal solubility and leaching [53,54], which can reduce accumulation and thus lower exceedance rates under comparable input conditions [52,53]. However, in certain regions like Rioja (63%), Catalonia (48%), and Castile-Leon (37%), exceedance rates in acidic soils were disproportionately high. This pattern likely reflects the overriding effect of elevated heavy metal concentrations in the sludge (Table 1) applied in these areas, which were large enough to offset the mitigating influence of low pH on metal accumulation. As such, heavy metal accumulation results from the interplay between soil pH and sludge characteristics and cannot be attributed to either factor alone.

Figure 3. Percentage of total plots that, after applying sewage sludge for 50 and 100 years, exceeded at least one heavy metal concentration limit with pH < 7 and pH ≥ 7. Grey regions (namely Asturias, Cantabria, Pais Vasco, Navarra, Aragon, Balearic Islands, Ceuta, Melilla and Canary Islands) did not register sludge application.

These findings indicate that, while sludge can be a valuable nutrient source, continuous application in certain regions may lead to heavy metal accumulation beyond safe limits, necessitating stricter localised regulatory oversight, pH management strategies, and periodic soil monitoring to ensure sustainable agricultural use.

We observed long-term shifts in heavy metal accumulation due to sustained sludge application (Figure 4), particularly emphasising the Zn and Cu exceedances after 50 years. These two elements are micronutrients, and thus essential for plant growth, yet their accumulation beyond legal thresholds raises concerns. At excessive concentrations, Zn can inhibit root development and microbial activity [55], while Cu can generate oxidative stress in plants, inactivate key enzymes for plant metabolism, and reduce soil biodiversity and consequently enzymatic activity in soil [56]. Although these metals pose relatively lower toxicity risks compared to other heavy metals (Cd, Hg and Pb), their build-up suggests that prolonged sludge application could gradually transition from nutrient enrichment to toxicity concerns. In addition, excessive Zn accumulation in soils can further reduce P availability by promoting the formation of insoluble Zn-phosphate complexes, which limit P solubility and uptake by plants, potentially exacerbating P deficiency and reducing P fertilisation efficiency in agricultural systems [57].

Figure 4. Distribution of heavy metal concentration limits exceeded after 50 and 100 years of P fertilisation by region, heavy metal and P source. The CaP scenario did not exceed any heavy metal limits in both time frames. Galicia and Valencian Community did not present any plots exceeding concentration limits in any scenario and time frame.

A more alarming trend emerges when considering the implications of sludge application over a century. In Murcia’s alkaline soils, Zn was the sole heavy metal exceeding limits in 100% of affected plots after 50 years. However, projections for 100 years of direct sewage sludge application reveal that Cd and Cu would also surpass regulatory limits, accounting for 18% and 4% of total plots exceeding limits, respectively. Given that Cd is highly toxic, bioaccumulative, and classified as a carcinogen, its increasing concentration in soil raises serious environmental and health concerns, as Cd can enter the food chain through plant uptake and groundwater via leaching [58]. Similar trends have been reported in previous studies, indicating that soils amended with sludge and mineral P can experience a build-up of carcinogenic heavy metals, whose leaching and/or plant uptake could lead to health issues and environmental degradation [59].

Our findings underscore the need for long-term sludge management strategies, particularly in alkaline regions where metals are more likely to accumulate than leach. While Zn and Cu are essential at trace levels, their gradual accumulation and the eventual increase in carcinogenic metals such as Cd necessitate periodic soil monitoring, stricter regulations, and possible amendments to mitigate risks. Without intervention, continued sludge application could lead to soil degradation, reduced agricultural productivity, and heightened risks of metal transfer to water sources, challenging the sustainability of sludge use in agriculture over extended periods.

3.1.2. Soil After Recycled Phosphorus Application

According to our model, the alternative application of struvite or CaP would lead to no heavy metal limit exceedance in any plot over the 50-year horizon (Figure 4). However, the use of vivianite would lead to Zn exceedance in one plot located in Castile-La Mancha. Zinc accumulation associated with vivianite application is due to sewage-sludge-derived vivianite, typically containing trace metals inherited from the sludge matrix. Consequently, even though vivianite is primarily composed of Fe and P, its Zn content can be higher than that of other recycled fertilisers and contribute to long-term soil accumulation when applied repeatedly.

Even after 100 years, no plots would exceed the limits by applying CaP, but in the case of struvite and vivianite, the number of plots would increase to one (in Castile-La Mancha) and 28 (three in Castile-La Mancha and 25 in Catalonia), respectively.

A considerably lower number of plots transgressed heavy metal limits when recycled P sources were compared to sludge. These results were expected, as the heavy metal concentration per kilogram of P present in the recycled products is much lower than in the sewage sludge used in Spanish cropland. Similar results were observed in previous reports [8,18], although in one study, the long-term heavy metal accumulation from direct sludge application was lower, mainly due to the difference in soil pH values [18].

Advanced P recycling also presents potential advantages in terms of lower input of contaminants, including PFAS and pharmaceuticals. While such pollutants are commonly found in sewage sludge, processes like crystallisation or thermochemical conversion involved in struvite and CaP production reduce their presence, improving the environmental safety of recycled fertilisers [60]. Furthermore, vivianite is especially promising for the Spanish context due to its compatibility with flooded paddy soils used in rice production [61]. However, the fertiliser value of vivianite is lower than that of struvite and CaP, mostly due to its reduced solubility under non-flooded conditions, limiting its efficiency to specific crops or soil types [12].

Regional variations in heavy metal accumulation are primarily driven by differences in sludge characteristics and amounts applied, as well as soil pH, influencing both metal mobility and accumulation trends. Regions with alkaline soils, such as Murcia and Rioja, demonstrated higher long-term heavy metal exceedances, particularly for Zn, Cu, and Cd, resulting from the continuous application of sewage sludge. Conversely, recycled P products such as struvite, vivianite, and CaP significantly reduce these exceedances due to their inherently lower metal contents. Consequently, regions applying sludge with higher pollutant loads or possessing alkaline soils incurred greater environmental externality costs associated with soil remediation and lost agricultural productivity. Ultimately, these regional differences in heavy metal accumulation directly shape the environmental externalities and influence the economic performance of recycled P fertilisers compared to direct sewage sludge application.

3.2. Economic Performance of Alternative Phosphorus Recycling in Comparison to Direct Sludge Application

While almost all NPVs across regions remained negative, the results indicate that some recycled P products performed relatively better than direct sewage sludge application, pointing to both economic and environmental advantages of P recycling technologies (Table 5). This relative advantage highlights the role of recycled P not necessarily as a commercially profitable venture, but as a public environmental service of reducing the pollutant burden of sewage sludge application while returning valuable nutrients to agriculture in a safer way.

Table 5. Net present value (NPV) in million EUR and benefit–cost (B/C) ratio of all P fertilisers after 50 and 100 years of application by region and treatment.

		Net Present Value (NPV) [Million EUR]		Benefit–Cost Ratio (B/C)
		50 years	100 years	50 years	100 years
Andalusia	Sludge	−1.12	−0.89	0.13	0.13
	Struvite	−2.04	−1.68	0.11	0.11
	Vivianite	−1.31	−1.14	0.25	0.24
	CaP	−3.60	−3.29	0.12	0.10
Castile-La Mancha	Sludge	−344.54	−268.69	0.11	0.11
	Struvite	−530.38	−437.57	0.13	0.12
	Vivianite	−334.21	−293.67	0.29	0.27
	CaP	−941.15	−863.19	0.13	0.12
Castile-Leon	Sludge	−67.91	−33.15	0.05	0.08
	Struvite	−29.63	−24.20	0.11	0.11
	Vivianite	−13.46	−11.81	0.29	0.27
	CaP	−41.98	−37.91	0.12	0.11
Catalonia	Sludge	−120.51	−88.54	0.24	0.25
	Struvite	−205.75	−168.37	0.32	0.31
	Vivianite	50.29	31.89	1.35	1.26
	CaP	−170.30	−169.39	0.56	0.50
Extremadura	Sludge	−111.04	−84.22	0.07	0.08
	Struvite	−156.19	−129.20	0.06	0.06
	Vivianite	−126.45	−109.43	0.14	0.13
	CaP	−308.54	−281.74	0.07	0.06
Galicia	Sludge	−1.60	−1.26	0.00	0.00
	Struvite	−1.34	−1.12	0.00	0.00
	Vivianite	−1.58	−1.35	0.00	0.00
	CaP	−3.35	−3.06	0.00	0.00
Madrid	Sludge	−10.33	−8.16	0.11	0.11
	Struvite	−18.39	−15.22	0.13	0.13
	Vivianite	−12.07	−10.65	0.32	0.29
	CaP	−34.02	−31.38	0.15	0.13
Murcia	Sludge	−63.33	−49.33	0.14	0.14
	Struvite	−114.43	−94.48	0.16	0.16
	Vivianite	−59.39	−53.36	0.43	0.40
	CaP	−192.80	−178.60	0.20	0.18
Rioja	Sludge	−60.13	−45.79	0.06	0.06
	Struvite	−84.10	−69.77	0.08	0.07
	Vivianite	−72.88	−63.10	0.15	0.14
	CaP	−175.88	−160.96	0.07	0.06
Valencian Community	Sludge	−0.45	−0.35	0.04	0.04
	Struvite	−0.66	−0.55	0.06	0.06
	Vivianite	−0.64	−0.56	0.12	0.11
	CaP	−1.49	−1.37	0.06	0.05

In Catalonia, vivianite achieved a positive NPV in both the 50- and 100-year horizons. This benefit stems from vivianite’s high P recovery efficiency, which can capture up to 60% of the influent P in WWTPs and provide high P surplus [62], while reducing the externalities associated with heavy metal accumulation. In most other regions, however, vivianite and CaP still reduced environmental costs relative to sludge, but did not reach positive NPVs, underlining the strong influence of regional conditions such as soil pH, pollutant loads, and sludge application rates.

In comparison, struvite demonstrates environmental benefits due to its low heavy metal content, making it a safer option in terms of soil contamination. However, its low P recovery efficiency (20–30% of influent P from WWTP) and relatively high treatment costs limit its cost-efficiency per unit of recycled P when compared to vivianite and CaP. Despite this, we observe an exception in smaller-scale regions such as Galicia and Rioja, where the NPV of struvite was slightly higher than that of CaP. This is attributable to a smaller scale of WWTPs that supply P demand. In such cases, the higher specific cost of struvite (EUR t⁻¹·P recovered) became less penalising due to the smaller absolute volumes involved. These findings imply that struvite may be a favourable option for smaller regions prioritising environmental safety over cost-effectiveness.

In contrast, CaP, which has the highest P recovery efficiency (up to 80% of influent P from WWTP), showed superior environmental performance due to low heavy metal content, eliminating indirect costs associated with soil remediation and cost of opportunity. Nevertheless, its economic appeal is strongly reduced by the high CAPEX, which dominated the cost structure over both 50- and 100-year horizons. Interestingly, while CaP had the lowest NPV in several regions over a 50-year period, its benefit–cost ratio remained relatively stable. This discrepancy reflects how the B/C ratio captures proportional efficiency rather than cumulative profit. Since CAPEX in the model was not included as a one-time investment but amortised annually, its impact was spread more evenly over time, which explains why the B/C ratio was higher than that of struvite despite lower NPVs. Similar observations were made in prior assessments of P recovery technologies across the EU, which found that high-CAPEX processes, such as CaP, may require regional subsidies to ensure long-term economic viability [9]. These results confirm that CaP can deliver long-term sustainability advantages if policy mechanisms such as infrastructure subsidies or pollution-based levies are implemented to compensate for the high initial investments.

An overarching trend across all recycled P products is that economic performance improves only marginally over time. With recycled P, B/C ratios remained relatively stable between the 50- and 100-year marks, reflecting how annualised CAPEX dominates the cost structure and how avoided externalities accumulate slowly. This contrasts with direct sludge application, where remediation and opportunity costs grew substantially in the long term. These results indicate that while advanced P recycling reduces environmental risks, its economic viability depends strongly on regional conditions and policy support to overcome the cost barrier.

The sensitivity analysis reveals how discount rates influence the economic performance of the studied P fertilisers. In the case of Catalonia for an application period of 50 years, the estimated NPV became less negative with higher discount rates for struvite and sludge (Figure 5). This is a trend observed in NPVs across regions mainly due to less weight in future costs compared to present costs at higher discount rates (see Equations (8) and (9)). In contrast, the use of vivianite in Catalonia began with a positive NPV at lower discount rates and diminished with higher rates as future benefits are devalued more steeply. This trend is comparable to typical NPV behaviour for increasing discount rates [63]. Meanwhile, the B/C ratio remained unchanged for struvite and vivianite, indicating a proportional evolution of benefits and costs that neutralised the impact of the discount factor. In the case of sludge, the B/C ratio improved with increasing discount rates, likely due to the delayed emergence of externalities such as soil remediation. Conversely, the high CAPEX of CaP implementation was penalised heavily in short-term periods, thereby decreasing the B/C ratio with increasing discount rates. This underlines how temporal cost structures and externalities affect the sensitivity of investment metrics in P recycling pathways.

Figure 5. Sensitivity analysis of net present value (million EUR) and benefit–cost ratio for different discount rates for Catalonia in a time span of 50 years by treatment.

Our findings suggest that in some specific cases, P recycling technologies provide clear net benefits over direct sludge application, though their efficiency depends on technological recovery efficiency from sewage sludge, CAPEX, and long-term soil impact. Vivianite appears to be the most balanced option, offering both high recovery and low environmental externalities, although its agronomic use is limited by low solubility under non-flooded conditions. Struvite is preferable in small-scale or highly sensitive regions due to its safety profile, whereas CaP provides long-term sustainability advantages, but requires higher investments.

These insights support policy efforts aimed at accelerating the adoption of nutrient recovery technologies, especially in regions like Catalonia where heavy metal risks and plot prevalence are high. Still, marginal differences in NPV and B/C among P fertilisers (e.g., Valencian Community and Galicia) highlight the need to assess overlooked factors such as water pollution, soil degradation, and organic pollutants like PFAS and pharmaceuticals. Potential benefits such as additional nutrients and organic carbon soil input should also be considered to fully support decision-making. Incentivising such transitions through infrastructure subsidies, pollution-based taxes, or regulatory frameworks aligned with the Soil Health Law could ensure long-term agronomic viability and economic resilience in Spanish agriculture [27].

4. Conclusions

This study provides an integrated assessment of the long-term environmental and economic implications of three advanced P recycling technologies (struvite, vivianite and CaP) compared to direct sewage sludge application in Spanish agricultural soils. By combining spatial soil data with a 50- and 100-year heavy metal accumulation model and cost–benefit analysis, we demonstrate that advanced P recycling can offer a potentially economically efficient pathway for mitigating heavy metal contamination while enhancing economic efficiency over time.

The results confirm that direct sludge application can lead to significant heavy metal accumulation, particularly of Zn, Cu, and Cd, posing risks to soil health, crop safety, and groundwater quality, especially in alkaline soils. In contrast, the recycled P products studied substantially reduce exceedance risks. Vivianite emerged as the most balanced option, combining relatively lower infrastructure costs compared to CaP, high P recovery efficiency compared to struvite, and low environmental externalities, though it has limited agronomic value.

CaP, in turn, requires higher capital expenditure but delivers superior long-term sustainability due to its negligible pollutant content and high recovery potential. Struvite, despite lower recovery efficiency, remains attractive for regions with small-scale agriculture or heightened environmental sensitivity. These environmental advantages are reinforced by the economic analysis. When accounting for indirect costs such as remediation and opportunity cost, both the economic net present value and benefit–cost ratios for recycled P options outperformed the baseline sludge scenario in some regions.

While our findings offer valuable insights, there are limitations, including assumptions of constant sludge application rates and the exclusion of metal leaching, crop uptake variability, and the potential agronomic benefits of organic matter and secondary nutrients. Additionally, the effects of emerging contaminants and broader environmental externalities remain insufficiently addressed, highlighting the need for further research. Despite limitations, this study presents an improved framework to integrate environmental externalities in the economic assessment to support sustainable agriculture while safeguarding long-term soil health at a regional scale. Future studies should refine these models by incorporating pollutant fate dynamics, crop-specific risks, and comparative life cycle assessments to inform more holistic nutrient management policies.