Chapter 4: La Bièvre: Evaluating ecological functions of Parisian plans

Recent restoration efforts along the Bièvre and the regulatory context in which they reside were integral in giving rise to current plans to recreate the stream in Paris. As stated in the objectives of the Mixed Syndicate and the Union of the Associations for the Bièvre’s Renaissance, the restoration of the Bièvre in Paris is a priority. In light of the difficulty in reworking the hydrological system upstream, the stream’s restoration is Paris was seen as a potential catalyst for the immediate suburbs. Supported by the Regional Counsel, several studies were carried out by its research department at the (IAURIF). In 2000, former Paris Mayor Tiberi supported the project and requested a study from the city’s planning agency, the Atelier Parisien d’Urbanisme (APUR). This chapter will examine this plan in light of restoration efforts upstream and then introduce the Institut d’Aménagement et d’Urbanisme de la Région Ile-de-France’s study on the feasibility of the Bièvre’s restoration in Paris. Based on the ecological functions presented in Chapter One, the success of these approaches will be measured.

APUR’s plan calls for the creation of three reaches of the Bièvre in Paris’s 5th and 13th arrondissements: within the Park Kellermann and Square René-Le Gall and along two streets near the Museum of Natural History. The project proposes the partial restitution of the Bièvre with two objectives. The first, labeled "environmental," would allow Bièvre water to empty into the Seine in a "natural" way instead of being sent to the Achères purification station as it is at present. The second, termed as an "urbanistic, landscape and cultural objective," would create in Paris a new "natural" and landscaped path at the scale of the promenade plantée of Bastille-Vincennes, destined to be prolonged outside of Paris to the Bièvre’s source. According to the plan, this course could be enhanced with a pedagogical presentation of the history of the places along it while simultaneously serving as a bike path.

The APUR pan explains that the modifications made in the nineteenth century to the geography of the capital permit only a partial open-air restitution of the Bièvre within Paris intra muros. Embankments and building construction realized in the 13^th arrondissement buried the Bièvre’s traces beneath them, as described in Chapter Two. APUR, consequently, chose three sites owned by the city where the Bièvre could be uncovered. The most space is available to recreate the stream on these sites and the ground level has remained reasonably unchanged since the nineteenth century, with little or no embanking having occurred.

In total, 2300 meters of the Bièvre’s former traces will be exposed to daylight in two different treatment types. The first is a sort of canal treatment with fencing surrounds that would be applied to sites where little space is available, such as along streets. The second treatment is a more "natural" type with vegetation-covered stream banks that could be realized within the two parks and on the Museum’s property.

The plan proposes a thread to complement these sites with a promenade linking them along the buried traces of the stream. This thread would take the form of a two-lane continuous bicycle path which would connect the Bièvre’s source to its confluence with the Seine. This cultural and recreational route envision was foreseen to include a redevelopment of adjacent public spaces (via limitation of automobile traffic and parking, tree plantings, etc.) and to be accompanied by an interpretative signage program and public welcoming spots with a small museum, bike rentals, etc. As these propositions have strong local implications, the plan suggests that they be made the object of large public collaboration. The modifications of public space involving the narrowing of streets and reduction of parking make room for the new bicycle path are in coherence with the politics of reducing traffic in the city.

The project proposes the restitution of two reaches of the Bièvre within the Park Kellermann which correspond to the stream’s formerly divided course. The combined Bièvre morte (dead) and Bièvre vive (alive) will measure 800 meters. The stream will emerge from waterfalls in a cave near the peripheral highway and immediately divide into two. Both branches will have vegetation-covered banks within the park, but one will turn into a steep-sided canal when outside the park. The square René Le Gall will also see the restitution of both reaches of the Bièvre, though rather than traversing the park, the reaches will border it, following a path similar to their historic course.

The stream will run as a canal beside the historic buildings of the Manufacture des Gobelins along the rue Berbier du Mets, where a new streambed will be built upon the collector which replaced the old streambed. Next, the plan proposes the redevelopment of the Bièvre’s traces within the Museum of Natural History’s annexes where its daylighting could "restore a veritable ecosystem" along a 400-meter stretch.

The last segment of the plan proposes the creation of a new outlet for the Bièvre next to the Pont d’Austerlitz near the historical confluence of the Bièvre and the Seine. APUR considers this an opportunity to develop a portion of the Seine’s banks more naturally than elsewhere.

The plan includes a hydraulic feasibility study that details potential solutions to supplying the Paris reaches with water from the Bièvre upstream. Based on the previous decision identifying the location of the new reaches, it examines several variants which were compared by a public water management agency. The feasibility study concluded that the supply of each reach must permit water flow at a speed of 10cm/s in an artificial bed of three meters in width by 0.5 meters in depth, which necessitates a rate of flow of 150 l/s. A supply that is at least partially water from the Bièvre was used as a base hypothesis, as this criterion was made explicit in the project objectives. It was supposed that the rate of flow of the stream at the entrance to Paris would remain identical to the actual rate of flow of the Bièvre within the deversoir, which is between 600 and 700 l/s in winter (outside flood periods) and only 200 l/s or less in summer. The rate of flow necessary at the head of the new system is only 200 l/s, which can be provided by the Bièvre deversoir. Water loss due to leaks and evaporation were calculated at fifteen l/s per reach, supposing the streambed would be concrete or stone-lined to provide good waterproof quality.

APUR illustrates how water from the Bièvre deversoir will be captured at the entrance of Paris and transported to the Seine via various conduits. Bièvre water will be first treated in a purification system beneath the Park Kellermann. According to SAGEP, this treatment is necessary to ensure water quality that is clear, healthy for public contact and marine life and maintainable during rainy periods. For each of the redeveloped sites, a system to supply the reaches with recycled water will be put in place. Purified Bièvre water will be transported inside pipes within existing collectors from the treatment plant to each of the reaches. Water within each reach would then be recycled and supplemented by limited amounts of replacement water (15 l/s per reach for compensating loss). A pipe placed in the deversoir would supply the outlet into the Seine and permit a residual rate of flow of 15 l/s. As the cost of the water treatment station represents 37% of the total investment in the project, reduction in its size or altogether removal would be advantageous. As described in the last chapter however, the quality of Bièvre water upstream must first be made satisfactory for alimenting public parks.

The three other variants were rejected for various reasons. The first involved creating a continuous flow of the Bièvre’s water from the entrance to Park Kellermann to the Seine with no recycling involved. Its result was considered to be identical to the retained solution with a 30% greater cost and a required flow of 400 l/s, which would not be available in the summer months without tapping the Seine for non-potable water. The other two solutions involved varying combinations of Bièvre water and non-potable water pumped from other sources. They did not correspond to the objectives laid out at the beginning of the study or to the demand of the associations.

The Absence of Ecological Functions

This project essentially aims to recreate portions of the Bièvre along its historic traces in a very modest and reserved manner, all while calling the project a "renaissance". While it responds to the associations’ objective of creating an opportunity for recreation and exhibiting the stream’s cultural significance, it pays very little attention to the ecological functions that characterize streams and fails to insinuate that the Bièvre will be reborn. First and foremost, the plan does not establish a continuous flow from the entrance to Park Kellermann to the Seine. APUR’s own environmental objective of "allowing water to empty into the Seine in a natural way" is not achieved. Although water from the Bièvre will be used to supply each reach, this water will not flow continuously but rather circulate within its respective area before being emptied into the sewer system. This would restrict marine habitat to certain reaches. While the vegetation-covered banks combined with clean flow of water may allow some life forms to persist, very little species movement will be possible.

APUR’s attention to the stream’s hydrological cycle is equally deficient. A permeable stream bed would greater serve the stream’s hydrological cycle than a concrete or stone-lined one by allowing water to be stored in the ground. Water from the surroundings could also be diverted into the stream. In the Swiss Brook Concept, water is collected from roof runoff and routed into streams. The intention to create vegetation-lined stream banks may serve as pollutant filtration and erosion prevention, though neither function is detailed in APUR’s plan.

Several challenges are outside of the scope of this individual project, as they concern the quality of the Bièvre water as it enters Paris. The potential to reconnect the stream to its watershed is difficult for this reason, requiring the opening of the Bièvre immediately upstream. Without sufficient water quality upstream, the integration of a purification station into Park Kellermann is inevitable. This raises the issue whether it is not more logical to wait for water to be cleaned upstream before carrying out a project in Paris. While Chapter Three highlighted the restoration efforts underway, the Paris project may have to wait to avoid the construction of a purification plant under Park Kellermann. This will be explored in Chapter Five.

The Ile-de-France region has been involved in the restoration of the Bièvre in collaboration with the communes it crosses and the different water management and development agencies. With these partners, it founded the association Bièvre Rivière Ile-de-France in June 2000. The project to restore the stream is a regional one that engages numerous partners which pose very contrasting problems, such as the preservation of the upstream course, the reopening of the suburban portions, and the restoration of the Parisian segment. The city of Paris, via the Atelier Parisien d’Urbanisme, identified three sites where the Bièvre could be reopened in the city and proposed a hydraulic system using water from the Bièvre, which was detailed in the earlier part of this chapter. Simultaneously IAURIF prepared a study of the entire Bièvre valley in 1999 and a study on the upstream portion. Most recently, it prepared a study comprising a historic atlas of the Bièvre in Paris and a presentation of development principles for the Bièvre’s Parisian traces, whose propositions reveal much greater attention to restoring ecological functions to the Bièvre.

Created in 2003 by landscape architect Pierre-Marie Tricaud, IAURIF’s study takes a comprehensive approach to the Bièvre’s historic traces and potential restoration. Its introductory text frames the problem and what is at stake in any urban stream restoration, with analysis of various restoration approaches and their relation to nature. Principles and solutions pertinent to the Bièvre in Paris and applicable to numerous other urban stream restorations are then proposed. The study saves its descriptions for redevelopment of the Bièvre in Paris for last, which include, among others, the three sites identified by APUR. The project’s objectives involve the domains of nature and history consistent with the will of the associations and those outlined by APUR. The desire to restore the stream entails both recreating natural milieu within the urban environment as well as evoking its history.

The IAURIF plan first sets out to identify the characteristics that make a stream natural and living. It breaks these characteristics into four categories pertaining to the configuration of the stream bed, the stream flow, and the degree of the stream’s openness to air. The lateral variation of stream bed ranges from a meandering form to a canalized bed. The vertical variation of streambed concerns its height and the rate of water flow and ranges from very variable to controlled, constantly in descent, or constantly horizontal. The stream flow can be gravitational with or without falls or involve a system of siphons to direct it. The stream opening can be open-air, encased in steep sides, covered under pressure (in a pipe) or covered and free-flowing (with an air pocket at the surface). The following chart details each of these characteristics with respect to their degree of naturalness and artificiality.

This plan analyzes and classifies the degree to which each of the preceding factors permits a natural and living stream. Natural is loosely defined as that which is reconstituted by humans to resemble milieu outside of human influence; living is defined with biodiversity of all different plant and animal species as the principal indicator. The table shows that for the factors examined, the natural and living character of the stream have a direct correlation. The accompanying text emphasizes that the artificial factors are not completely inert, or lifeless (such as a covered stream with air pocket). Creating a living stream is less difficult than creating a natural one, especially in an urban environment.

The previous examination determined the most natural and living features suitable and attainable along the Bièvre’s Parisian course. While reestablishing a meander was deemed impossible, certain areas of the stream could have wide banks and thus the possibility to overflow. With regard to vertical variation, some flexibility in water level would be feasible with controlled flooding. Concerning flow, gravitational flow was found essential to assure biological continuity. Several other factors were analyzed in the plan, such as whether the streambed should be reestablished in the exact location that it was buried. The stream bed’s location with respect to ground level was expected to vary in relation to its former position. The anticipation of good water quality, which can move slowly, allows for a very slight descent in the stream’s course. This avoids raising the water level by pump, as proposed in the APUR plan. With regard to the stream’s opening, several types of course were envisioned: an accessible stream (the most free-flowing), steep-sided and subterranean (the most restricted). Among the features studied, the following were found applicable and desirable for the project: a fixed or canalized stream bed, a controlled rate of flow and height, gravitational flow, and an open-air or unpressurized underground course. These were determined compatible with each other and with the desire to establish gravitational flow.

The Institute d’Aménagement et d’Urbanisme de la Région Ile-de-France worked from the Atelier Parisien d’Urbanisme’s hydraulic plans and a list of criteria formulated by the Union Renaissance de la Bièvre to establish continuous Bièvre water flow from the collector at the city’s periphery to the Seine. The Union’s proposal called for the use of Bièvre water alone, but the interruption of continuity with the passage by pressurized canalization was seen to severely (as seen in the APUR plan) compromise the stream’s biological potential. For this reason IAURIF went one step further in its plan by replacing the canalized passages by underground channels at atmospheric pressure. Tricaud criticized the APUR plan in stating that "its solutions can not in any case pretend to constitute a restoration of the Bièvre, not even a recreation of the stream." As restoration implies bringing back a former state and a recreation describes making something anew, Tricaud’s criticism is accurate. The APUR plan does neither of these things. For these reasons, the IAURIF plan concludes that an entirely gravitational and continuous flow is essential in order to call the project a restoration of the Bièvre. The IAURIF approach claims that is more respectful of the Atelier Parisien d’Urbanisme’s objective of "allowing water to empty into the Seine in a natural way" than APUR is itself. The IAURIF plan states that an uncovering of certain segments separated by canalizations (as proposed by APUR) may constitute a primary stage, but only if a [complete] renaissance of the stream will be realized later.

The IAURIF plan’s objective to create a gravitational flow imposes a uniformly descending profile that constrains the stream bed’s location. The water level within the stream along its course must therefore be carefully considered. The Bièvre will be open-air and steep-sided for a large portion of its course. The open-air quality will allow for the oxygenation of the water and thus promote cleaner, purer and more habitable water. This configuration has a practical advantage in that the wet section (when not flooding) will be larger than that of a shallow-dug stream or that of an underground conduit. Its capacity to stock flood waters is therefore larger.

The plan details three potential stream configurations based on the depth and width of the stream opening. The first, labeled the "accessible stream", has an opening that permits stream banks of an angle less than forty-five degree from the ground level on each side. The opening, or width of available space, is at least equal to the width of the stream plus the double of the depth. As he stream is often three meters wide along these stretches, the opening must be seven meters wide with a stream depth of two meters. A stream at two meters deep requires seven meters of available space in order to have forty-five degree angle banks. The second configuration, labeled the "steep-sided stream" is created when the opening does not permit stream banks of less than forty-five degrees but allows creation of a trench that is not deeper than it is large. This occurs when the width of the opening is less than twice and greater than half of the depth. The third configuration is the "subterranean stream," which occurs when the opening does not permit a trench as described above. This occurs when the width of the opening (minus stream width) is less than half the depth. It is completely covered, but its unrestrained nature permits normal rate of water flows and oxygenation.

The "accessible stream" can be created when the stream bed is close to the actual ground level and where the opening is sufficient in size. In the Park Kellermann, however, where the required dimensions for the Bièvre’s gravitational flow puts the stream bed 3 meters below the ground, it is possible to run the Bièvre through a rocky gorge in lieu of creating a 9 meter-wide valley. Outside of the three parks, there is hardly any public space large enough to accommodate the accessible configuration, though two places could possibly become available in the future after changes in the urban fabric occur. Within the parks, the stream can be developed in a rather natural and living character in harmony with the gardens crossed: low accessible stream banks and riparian vegetation.

The steep-sided stream is the result of the reduced size of the opening. The plan calls this not inconvenient, but practical. As most sites of this type are in the street, the steep-sided nature and consequent inaccessibility of the site will shelter the stream and its lively milieu from agitation and risk. The passage of the stream between walls will not prevent the presence of aquatic vegetation, bird nesting spots, etc. One side can be very urban and the other very natural.

Subterranean sections of the stream can also have natural, authentic and attractive elements. The essential thing is that the section is sufficiently sized to assure a free flow while maintaining a pocket of air permitting the oxygenation of the water. The APUR plan proposes a series of conduits to link the sections of the Bièvre that would be open to the air. This plan on the other hand proposes using old vaults created for the Bièvre eighteenth and nineteenth century, among others. Oversized for their role as local sewers, they were abandoned for conventional sewer pipes. These vaulted areas would provide ample space for the Bièvre and an air pocket as well as permit natural light to pass through openings, or oculus zenithal.

The IAURIF plan provides ample evidence of the feasibility of reestablishing an ecologically-functional stream, but also commits to justifying the reasons for doing so. Recent attention has been paid to increasing the amount of green space in the city with the creation of three parks and several new promenades linking the center of the city the countryside. However, there are still several neighborhoods with no links to green spaces. Creating them is difficult with little space available, but the Bièvre’s traces present one possible opportunity. These traces could be used to link three small parks to create a "park-promenade" or "park-stream". It could become a privileged place for recreation, fishing, etc. If not completely uncovered, it could still become a favorable promenade. Among others, the plan refers to Olmsted’s Emerald Necklace as a precedent.

A Model of Ecological Engineering

This plan does not conceal its emphasis on water treatment using natural elements: ground soil filtering, oxygenation by waterfall, and decantation by macrophytes within retention basins are all described. The focus is on treating stormwater, not wastewater, and the interest is as described partly pedagogic: ecological purification devices in the center of Paris would allow many people to understand how the system functions and would promote it for other sites. Certain steep-sided spaces could be enlarged to contain retention basins for decantation, or separating water from sediment deposits. Rainwater would be received and treated here before being released in to the Bièvre. In the gardens such as René-le-Gall, small decanting basins could be planted with reeds. While road run-off is filled with pollutants requiring several stages of purification, roof run-off is much cleaner and could be stocked in cisterns and used to irrigate vegetation alongside the stream.

The portions of the Bièvre located within steep-sided trenches could present a great advantage in terms of flood risks, as the steep sides would allow the water level to rise securely.

This flexibility of this configuration outperforms conduits in the case of rising water level. The accessible stream is even more advantageous, as it offers a larger wet section than the conduit and therefore a greater capacity to hold water. All floodable spaces situated between the habitual level of the water and the ground level could become areas where flood waters could expand and should be developed with this in mind. These spaces can even be creative, making places like the Square René-le-Gall itself a possible retention basin. The interest in carrying out such projects is again pedagogic, as they would not absorb a flood in its entirety, but serve to show that the Bièvre could overflow without being dangerous.

This plan claims that the complete restoration of the Bièvre will permit biological continuity thanks to 1) the continuity, assured by gravitational flow without significant falls; 2) the oxygenation, assured by the course that is mostly open-air or covered with an air pocket; 3) the light, assured by the predominantly open-air course. According to the plan, the continuity of natural milieu is essential to biodiversity, as it permits species to occupy different areas responding to their needs at different stages in their lives. It ensures the intermingling of species of animals and plants which results in greater fertilization and dispersal. The liaisons assuring this continuity are called biological (wildlife) corridors. The continuity of airborne milieu can permit minor interruptions, as the majority of animal species, even terrestrial, cross water courses and inert spaces. These do not present challenges for pollen and seeds, which are transported by wind, birds and insects. The continuity of aquatic milieu is more difficult to insure: in a forest-like environment, ponds are linked by humid or shaded areas which permit the migration of batrachians, but these conditions are not easily found in cities. As for entirely aquatic species, an entirely aquatic continuity is essential, without crossing obstacles, with a sufficient oxygen level along the entire course, and nearby feeding areas. The water courses are habitually the principal vectors of this continuity when they respect these conditions.

The IAURIF plan calls for the establishment of a continuous course of Bièvre water from the entrance to Paris to the Seine. Various methods are used to maintain a gravitational flow of clean water throughout this course, including both above and below ground segments. The scheme calls for the development of the three sites outlined in the APUR plan, as well as several others. For the Park Kellermann, the recreation of only one reach is proposed because the other (Bièvre morte) would be too constrained by its surroundings. The Bièvre vive would be developed very close to its original stream bed (one to two meters above it instead of three to five meters). The stream’s depth would oblige it to cross the park in a steep-sided valley where man-made rocks could create an ambiance reflecting a gorge.

Other portions would be created where the stream would pass underneath the street in a cave-like space. Water would not be subjected to pressure here, as air and sometimes light would enter from above. Other portions would involve housing developments which could help finance the cost of the stream’s restoration. This portion at the Place de Rungis does not correspond to the Bièvre’s historical traces, yet provides the only open space to recreate the streambed. The Bièvre’s opening is also described within a 1960s housing superblock, îlot Bièvre (discussed in Chapter Two, page thirty-nine), which would attempt to reverse some of the most recent destruction of the stream. The permission of property owners is the greatest challenge here and for several along other sites the path.

The Square René-Le-Gall and Rue Berbier-du-Mets are two sites targeted in the IAURIF plan as in the APUR plan. Like Park Kellermann, however, they are modified to account for the gravitational flow, which influences the level at which the stream bed is dug out, and greater attention to ecological functions. This square offers a unique opportunity as the site where the ground level is closest to the stream’s original bed. Adjacent to the square is the Rue Berbier-du-Mets, one of the streets which follows the Bièvre’s former course (Bièvre vive). Here the original streambed and bank still exist, covered with a simple vault from 1864 to 1897 for use as a local sewer. While the APUR plan calls for the creation of a new streambed above this vault, the IAURIF plan proposes to uncover the original streambed.

The stream’s restoration within the annexes of the Museum of Natural History is more developed than that outlined in the APUR plan. Here the remarkable and spontaneous nature of the trees and plants along an alley closed to the public giving have potential to create another "accessible" stream. From this section the Bièvre runs through the Jardin des Plants before entering the Seine. As this large open space provides ample opportunity for the Bièvre’s most unconstrained form, it outperforms APUR’s option which ran the stream through more confined spaces in order to keep it closer to its historic course. This is congruent with other parts of the plan, which are less attentive to following the Bièvre’s traces exactly than to recreating an urban stream. He justifies this with the fact that stream courses shift over time and therefore do not need to be maintained within one of the streambeds that they occupied in time.

Integration of Ecological Functions

The plan of the Institut d’Aménagement et d’Urbanisme de la Région Ile-de-France represents an ecological approach to reestablishing a stream in an urban environment. Despite the absence floodplain along the majority of its course, the stream’s configuration allows it to successfully carry out numerous ecological functions. It meets the criteria established in Chapter One that a stream should be: