3.1. Case Study—Chełmońskiego Street Housing Estate in Cracow
The analyzed area is located in the northwestern part of Cracow, District IV Prądnik Biały, between Chełmońskiego and Piaskowa Streets. The nearby area is characteristically made up of low-rise buildings. villas, terraced, and single-family houses. The northern neighborhood has green areas and the stream of Sudół. The satellite image below (
Figure 1) presents a detailed area of the analysis based on which it is possible to determine the type and character of its development.
The analysis encompasses an urban area currently residential in nature. Within this development, 16 multi-family buildings of low-rise character (4-story buildings) were completed, with a total of 275 flats. The buildings are situated on both sides of the estate road, which forms the compositional axis of the analyzed neighborhood. They differ in shape and surface areas depending on the location in relation to the road. The central part of the estate features a car park dividing the estate into two symmetrical parts. The remaining on-street parking spaces were designed alongside the road in front of the buildings. Overall, there were 123 on-street parking spaces and 196 underground garage spaces allocated for the use of the residents. There are two children’s playgrounds in the northeastern and southeastern parts, both surrounded by green lawns and low-growing shrubs. However, there is a lack of green areas for recreation and sports for adults or any form of a green park for the residents. The area is also flat with no hills or stairs. The surface area of the analysis is over 2.4 ha, the detailed land use is presented in
Table 2 [
78].
The estate was fitted out using impermeable materials for the paved surfaces of the vehicle and pedestrian traffic zones. Cobblestone was used for the estate road, on-street parking spaces, pavements, and access passages. Permeable areas are only green areas made up of small lawns near the parking spaces and the fence, as well as backyards of the ground floor flats. There are only a few trees growing in the whole estate, the dominant greenery is low-growing shrubs and ornamental bushes. The estate development indicates a high degree of catchment surface sealing. Therefore, rainwater has a limited capacity of infiltrating the soil where it falls and, in consequence, contributing to the increased intensity of surface runoff.
3.3. Solutions
The selection of LIDs due to the existing buildings, land use, and limited space was made without optimization algorithms, based on a local vision and pragmatic assessment of the techniques that can be used. Each building has 4 downspouts, the green areas around the buildings are small and the 3 gutters have their mouths within private small gardens adjacent to the ground floor flats; therefore, for these gutters, soakaways are planned that do not take up a large area and would be possible even within these private gardens. At each building, one corner, and thus one downspout, is within a public green area, not belonging to any apartment. For such an area, a rain garden is planned for one downspouts from each building.
Among the existing types of impervious areas, changes are proposed in the discharge of rainwater for pavements, parking spaces, and car parking due to the ease and price of their implementation. Their surface is made of cobblestones (easy to dismantle) and has thinner sub-base compared to roads. On pavements and parking spaces, it is planned to replace the pavement and the sub-base layers, for the parking lot: dismantling of the surface area, installation of infiltration boxes, and rearrangement of cobblestone surface. No changes are planned for roads because the changes would be too expensive compared to the abovementioned areas, and there is no free space for retention of water collected from roads.
3.3.1. Pedestrian Streets of Mineral Surface
Mineral permeable surfaces on pavements allow the phenomenon of infiltration, i.e., the soaking of rainwater into the soil and groundwater. Implementing this type of solution requires soils of good and moderate permeability as well as distance of at least 1 m from the level of groundwater table [
79,
80]. Surfaces allowing for slow infiltration of rainwater prove to be a good solution in urban areas. Implementing them into the existing developments or slightly modifying current technical infrastructure elements does not require a lot of effort [
81].
In the entire estate, it is proposed to replace the surfaces of existing pedestrian routes made of cobblestones with naturally stabilized, environmentally friendly mineral surfaces.
As a result of the implementation of the aforementioned solution, rainwater infiltration deep into the ground will increase while alleviating the burden on the rainwater sewer system. This will lead to a decreased impermeable surface within the estate.
Table 6 presents, for comparison, values of average annual runoff from paved areas before and after the surface change.
3.3.2. Parking Spaces of Permeable Surface
In order to increase surface infiltration in urban areas, applied are surfaces made of materials or arranged in a way enabling water soaking in. The proposed solutions require soils of good and moderate permeability as well as a distance of at least 1 m from the level of groundwater table [
80]. It is proposed to use porous pavers, made up of specially designed chamber structures of high load resistance (even up to 350 t/m
2). They also have expansion gaps thanks to which the surface can “work” and the material does not deform or bulge even at extreme temperatures (−40 to +85 °C). High volume of chambers allows for plant growth as they optimize nutrient management, as well as air and water supply. Currently, on the market, there can be found pavers made of recycled synthetic materials.
By introducing the abovementioned solution, the area of impermeable surface will decrease eight-fold within the analyzed section (
Table 7). Implementing the pavers system in place of parking spaces will contribute to increasing infiltration deep into the ground and, consequently, decreasing runoff from the estate.
3.3.3. Rain Gardens
A solution combining rainwater management and the aesthetics is rainwater gardens. In terms of the visual, they should complement the landscape, however, their main aim is to reduce the surface runoff and increase infiltration excessive impurities removal (washed off from roofs, streets, car parks, and other impermeable surfaces) [
39,
82,
83,
84]. Following the subdivision mentioned earlier, rain gardens are an infiltration with retention type of solution [
85,
86]. The type of soil and the drainage area are key to determining the size and design of a rain garden. The depression can start at 1 m wide, 2–7 m long, and should be filled up with garden soil. It is necessary to leave around 15–30 cm of lowered space, which will store the runoff after for several hours.
Due to the close proximity of the residential buildings and the limited development area, 1 in-ground garden (so-called “wet garden”) is designed next to each house. It is located next to the outlet of one out of four drainpipes receiving water from the building’s roof. It is additionally equipped in an emergency overflow. Excess of water retained in the tank will be removed gravitationally by drainage to the designed soakaway, and in case of high volumes of water, to the rainwater sewer system.
This solution will enhance the biodiversity within the estate and improve the quality of life of its residents. Rain gardens will contribute to improving the quality of the drained rainwater by reducing surface runoff and, at the same time, reducing flood risk. An economic aspect of utilizing this solution is reducing the fees for non-managed rainwater and water services. Moreover, locating the rain gardens at the entrance to the buildings will enhance the neighborhood esthetics [
87].
It is proposed to design 16 rain gardens embedded in the ground, with a sealed bottom, each of 6.25 m
2 and in a shape of a rectangular measuring 2.5 × 2.5 m. They will be located at a 30 cm distance from the building’s façade directly under a gutter channeling rainwater from the roof to the garden. Lowering the ground level by approximately 30 cm, with the infiltration area of 6.25 m
2 offers a retention volume capacity of approximately 30 m
3. Each garden will receive water from approximately ¼ of the building’s roof, i.e., around 100 m
2. Design rainfall volume of intensity
Pmax (
t,
p) = 211 (dm
3/s∙ha) and duration 15 min for ¼ of surface area of 16 buildings is approximately 29 m
3 (
Table 8). Rain gardens are, therefore, able to harvest the design rainfall in full, and in case of higher volume, the excess will be drained to the sewer system.
3.3.4. Soakaways
One of the infiltration devices with underground retention is soakaways, the construction of which is similar to a simple concrete ring well. It has a permeable bottom made of a gravel layer through which the rainwater drains to the ground [
88]. Soakaways are used for local rainwater drainage to the soil that should be characterized by permeability coefficient above 10
–6 m/s. Soakaways are introduced in places where there is not enough space for surface devices [
89].
In order to drain the rainwater from the buildings’ roofs, it was proposed to implement 48 soakaways, concrete ring wells intended for surface water retention and draining it into the ground.
Height is assumed at 3 m, and diameter at approximately 1.5 m. A soakaway needs to reach from the ground surface to the permeable layer, and the distance between the lower ring and the groundwater level needs to be at least 1.5 m. The distance of the soakaway from the building should be at least 2 m. In order to improve infiltration, outlets of approximately 1–2 cm are applied to the walls of the lower ring.
Water will be channeled to the soakaway by drainpipes (positioned horizontally underground, below the frost depth) of full walls at 150 cm diameter. The outlet will be positioned 20 cm above the filtration layer. The soakaway will be filled with permeable material of washed sand (50 cm) and sandy gravel (100 cm). Each soakaway, similarly to the rain gardens, will receive water from approximately ¼ of the building’s roof, i.e., around 100 m
2. Design rainfall volume of intensity
Pmax (
t,
p) = 211 (dm
3/s∙ha) and duration 15 min for ¾ of surface area of 16 buildings is approximately 90 m
3 (
Table 9). The soakaways of the assumed measurements are able to harvest the design rainfall in full, and in case of higher volume, the excess will be drained to the sewer system.
3.3.5. Infiltration Boxes System
Infiltration boxes are devices made of polyolefin truss plates. They are recognized as infiltration devices with underground retention. Draining can only be done in permeable soil. The minimum depth of application should be 0.4 m for green areas. In case of higher intensity caused by vehicle traffic, the infiltration boxes should be arranged deeper by 0.8 m [
44]. Infiltration boxes can also be used in soils of permeability greater than 10
−6 m/s. In order to avoid the infiltration box’s capacity decrease due to sedimentation, it is necessary to carry out initial water treatment (removing suspended sediment). Thanks to the modular construction, it is possible to design objects of any size and adjust their spatial arrangement to fit in with the local conditions [
46].
On the car park’s surface (the central part of the analysis), a system of infiltration boxes was designed, aiming to manage rainwater from the car park’s surface by retention and non-pressure distribution and infiltration in the soil. Rainwater harvested from a hard surface will be drained off via linear drains, rain grooves, and cleaning devices for infiltration boxes. In order to design an infiltration box system adequately, the length L (m) was calculated based on the recommendations provided by one of the manufacturers in a formula [
90]:
where:
An—reduced surface area (m2), An = ∑(A∙ψ), ψ—runoff coefficient,
rd—rainfall intensity (dm3/s·ha);
D—rainfall duration (min);
b—infiltration box’s width (m);
h—infiltration box’s height (m);
sr—infiltration box’s accumulation coefficient;
kf—ground filtration coefficient (m/s)
Based on formula 6 and data presented in
Table 10, the length of infiltration boxes was calculated at 43 m. The set will be made up of 36 infiltration boxes measuring: 1.2 × 0.6 × 0.3 m. Their volume will be 7.6 m
3, which is enough to harvest the design rainfall volume from the car park (
Table 11).
3.4. The Effect of the Proposed Solutions on Rainwater Management
The key aim was to propose solutions to rainwater management at the existing housing estate. The proposed concept was intended to complement traditional rainwater drainage methods with alternatives facilitating the reduction of discharge to the rainwater sewer system and slowing down the runoff. The selected solutions can be implemented into the existing land developments where no space had been previously left for green areas. Within an estate, green areas occur mostly in form lawns around buildings with low-growing plants, ornamental bushes, while the remaining spaces are pavements, roads, parking spaces, and car parks, all hard-paved. It was proposed to resurface these areas and their profiles, on pavements and parking spaces, which would be cheaper than resurfacing roads. For the runoff from buildings’ roofs and the communal car park’s area, it was proposed to implement infiltration devices in forms possible to be adopted in the existing development. Within the estate of total surface area 24,400 m
2, the sealed surfaces (roofs and paving) take up 13,250 m
2, while the proposed green solutions are designed to take up nearly ¾ of this surface area, i.e., 9800 m
2 (
Figure 2), including:
resurfacing areas and their profiles:
pavements (2300 m2)
parking spaces (780 m2)
management of the rainwater from the roofs of residential buildings through introducing:
rain gardens (1600 m2)
soakaways (4800 m2)
infiltration of rainwater from the car park (470 m2) through implementing infiltration boxes (stormboxes).
For the proposed solutions’ effectiveness assessment runoff from particular areas was calculated. Subsequently, the surface runoff values from before and after the estate completion and after any potential implementation of “green” solutions were summarized. Cumulative results are presented in
Table 12 and
Table 13.
From the summary in
Table 12 transpires that the proposed solutions can considerably reduce surface runoff from the estate. This is thanks to the reduced sealing as a result of implementing permeable surfaces and solutions enhancing water infiltration deep into the ground.
3.5. Financial Effectiveness Assessment of the Proposed Solutions
A calculation of approximate investment costs that would need to be incurred in order to implement the proposed solutions. Cost estimation is based on local process available in catalogs by manufacturers providing the abovementioned materials and products. Additionally, prices from recently concluded tenders, which involved works within the same project scope, have been researched. All amounts were converted into the 2018 prices using the Consumer Price Index (CPI). The estimated investment costs are calculated as a unit cost multiplied by a quantity. They are summarized in
Table 14.
The cost of implementing the above-given solutions was estimated at € 169,600. It is worth noting that the cost of traditionally executed pavements and parking spaces is similar to the one proposed here; therefore, the only additional cost for the developer is made up of rain gardens, soakaways, and infiltration boxes at the total cost of € 79,635.
In relation to the prices of the properties, whose total surface area within the estate is 20,000 m
2 it can be calculated that the cost of the proposed solutions (rain gardens, soakaways, and infiltration boxes) per 1 m
2 is merely €4/m
2, and per household, with the total of 275 flats, is € 290/flat. This cost is low and does not enhance the developer’s profitability as it only makes up 0.2% of the average price, which was € 1643/m
2 in Cracow in 2018 [
91]. Despite the costs being so low, the developer did not plan or execute any similar rainwater retention solutions within the estate. This might have come as a result of the lack of such recommendations in terms of spatial development, a local spatial development plan was introduced merely a year after the estate had been completed. Currently, this plan includes a recommendation on local rainwater drainage through retention or drainage via rainwater sewer system, including solutions which facilitate rainwater drainage to soils, slow down discharge to receiver, and improve retention.
Additional incentive toward changing the investors’ attitudes may also be water services fees—fees for rainwater drainage, introduced in 2017 by the Water Act. Below is a calculation of an annual fee for rainwater drainage from a sealed surface to the rainwater sewer system to be incurred by the estate’s administrator. Fees were estimated for three calculation variants (
Table 15):
variant A—current development status, sealed surfaces, lack of rainwater retention solutions,
variant B—permeable paving and parking surfaces, lack of rainwater retention solutions,
variant C—permeable paving and parking surfaces, rain gardens, soakaways, and infiltration boxes.
As illustrated in
Table 15, if the developer or, rather, the later administrator of the estate, decided on rainwater drainage directly to a watercourse, they would pay on average € 1246 to € 1487 per year. The fees would be lower by € 241 if they made the paving and parking spaces using permeable surfaces (variant B), which would reduce annual surface runoff by approximately 1300 m
3. It needs to be noted here again that this would not increase the total cost of this estate’s development.
If the developer or administrator additionally introduced the proposed “green” solutions, the fee could even be reduced to € 415. A large reduction of € 831 results from two things, the surface runoff reduction (reducing discharge from roofs and car parks) as well as lower fee rate eligibility. The Water Act stipulates that the fee rates are reduced when retention devices are implemented; the reduction depends on the devices’ volume capacity that is related to the average annual runoff volume. In the proposed concept, the combined retention capacity of devices (gardens, soakaways, and infiltration boxes) is approximately 133 m3 that makes up for 2% of runoff from sealed surfaces, therefore, it is possible to apply a lower rate of € 0.15/m2.
In order to assess the financial profitability of investing in rainwater retention devices, a calculation was done to extract a simple investment capital payback period (
Table 16).
The estimated payback period, surely calculated and taken into account by investors and estate administrators, is unfavorable—return on investment would come after a period of nearly 100 years. Undoubtedly, one can expect that such a result will not be motivating enough to decide on expenditure of € 79,635 if the benefits in terms of the rate of return will be at 1% annually.
As mentioned in Chapter 2, many countries have a functioning system of municipal fees related to the particular way of land development and rainwater management, being a strong incentive toward investing in rainwater retention. Cracow has yet to introduce such fees, however, several Polish cities have introduced them already. Below is a calculation of a fee for rainwater drainage to the rainwater sewer system. Fees were estimated for three variants (
Table 17), assumed rates were € 1.17 and € 1.00/m
3, 15% lower rate for the implementation of retention devices (similarly to the national fees):
variant A—current development status, sealed surfaces, lack of rainwater retention solutions,
variant B—permeable paving and parking surfaces, lack of rainwater retention solutions,
variant C—permeable paving and parking surfaces, rain gardens, soakaways, and infiltration boxes, rate of € 1.00/m3.
Depending on land development and rainwater management, the administrator would pay on average between € 2769 and 9665 per year. Introducing permeable surfaces to pavements and parking spaces allows for a reduction of € 1565. Utilizing retention devices offers a saving of further € 5331, at the lower fee rate. The saving may go up to € 6896 at the lower rate, compared to the fees in variant A. Such benefits may prove to be a considerable incentive for deciding on an investment, the financial profitability indicators are summarized in
Table 18.
The estimated values of the financial profitability indicators are positive, the investment is paid back to the investor within a period of 12 years, and the average annual fee reduction is nearly 10% of investment costs. It seems that given an adequate information campaign, administrators and residents would be willing to participate in individual investments in terms of rainwater retention. Contrary to the financial effectiveness indicators calculated above and based on the benefits from reducing the water services fees (96 years and 1%), the calculated indicators resulting from the benefits offered by reducing urban fees (12 years and 9%) show values which prompt to making investment decisions. For example, “My electric current” program implemented by the government in 2019 was calculated based on a 10-year payback period and has attracted a lot of social attention (within two months, there were approximately 4000 applications submitted; the program involves a system of tax exemptions and subsidies to fund photovoltaic microinstallations). The stronger incentive of urban fees proven here, as compared to the water services fees, is a result of the rates taken for the analysis of the rates figures. In terms of other rates and their reductions given the usage of retention devices, their efficacy will naturally change.