PoroElastic Road SUrface: an innovation to Avoid Damages to the Environment

Project Information

PERSUADE

Grant agreement ID: 226313

Closed project

Start date 1 September 2009

End date 31 August 2015

Funded under

• Specific Programme "Cooperation": Environment (including Climate Change)

Total cost € 4 650 637,30

EU contribution € 3 376 692,30

Coordinated by

CENTRE DE RECHERCHES ROUTIERES - OPZOEKINGSCENTRUM VOOR DE WEGENBOUW
Belgium

Final Report Summary - PERSUADE (PoroElastic Road SUrface: an innovation to Avoid Damages to the Environment)

Executive Summary:
The aim of the PERSUADE project was to develop a poroelastic road surface (PERS) which offers extremely high noise reduction, and at the same time is safe, cost effective, sustainable and has an acceptable durability, and to demonstrate this on trafficked roads. For this, a holistic project plan was drafted, comprising the study and documentation of all relevant aspects of this novel type of road surfacing.
A stepwise approach has been followed comprising laboratory testing, small-scale pilot test tracks and full-scale test tracks on trafficked roads, which were then extensively monitored. In parallel, the environmental impact and the economical aspects (cost-benefit) have been studied. This report deals with all the technical findings of the project.
For the mix development, the team started with some target properties, such as durability, friction, noise absorption, elasticity and rolling resistance. Several PERS mixes were tested in the lab and it turned out not to be difficult to obtain reasonable values for the targets, except for durability. It was a real challenge to reach an acceptable ravelling resistance, but one of the major successes was that two PERS mixes with a laboratory ravelling resistance better than that of a high quality thin asphalt layer was found. One of these mixes had a lower rubber content, about 20 % by weight, and the other one had a higher rubber content, nearly 40 %. To meet the friction target was no problem; neither to obtain good sound absorption and sufficiently low stiffness.
The structural design of a road with PERS as wearing course has been studied and several experiments have been carried out. This is an important issue as one has to make sure that the road structure does not fail due to an expected limited bearing capacity of the top layer. The PERS must also not delaminate from the sub-layer. Although the team succeeded to bind the PERS layer sufficiently to the sub-layer initially, this turned out to be the major challenge.
Small-scale test sections on little or not trafficked areas were built, allowing contractors to practice with the novel technique and to monitor some crucial parameters, such as friction, durability and winter behaviour. The consortium built six of these pilot test tracks, instead of one as foreseen in the project plan, but they turned out to provide a lot of useful experience, and the team could then proceed with the construction of the full-scale test tracks on trafficked roads. In total, eight test tracks were constructed, using three different techniques: application of an in-situ made PERS mix (similar to the application of asphalt, but without heating the mix), gluing of prefabricated panels to the sub-layer, and gluing small “tiles” on cement concrete blocks which were then laid in the conventional way. These experiments showed that it is possible to achieve an acceptable result, but it is easy to make mistakes. Such mistakes may be working in a humid environment, failure to protect the roller efficiently from particles sticking to it, or letting the tack layer cure too long before the application of the PERS layer, as the system is not forgiving.
The monitoring time was unfortunately much shorter than originally foreseen (less than 1½ year instead of planned four years). This was due to delays in the construction of the test tracks mainly since the team had to work longer with developing durable mixes. Nevertheless, a lot of information was extracted from the test tracks.
The friction on the full-scale test tracks, measured with high-speed devices, was no problem, provided necessary pre-conditioning of the material was done in some cases. The noise reduction was impressive. Compared to an SMA 0/16 which is very common in Sweden, a noise reduction for cars between 8 and 12 dB(A) was obtained and compared to SMA 0/11, which is a more common reference in Europe, the noise reduction was found to be between 6 and 10 dB(A). The PERS mix with the higher rubber content yielded the highest noise reduction. Sound absorption was quite good, but varied significantly from test track to test track. The rolling resistance for car tyres on PERS was found to be of the same order as for SMA 0/16, hence 10 – 20 % higher than the best performing conventional pavements, but for truck tyres it was about 50 % higher. As PERS is only intended for application on a very limited fraction of the network (on hotspots), this will not influence significantly the fuel consumption and CO2 emission.
An important issue that has been studied is the winter behaviour and it turned out that it behaves differently under winter conditions: the surface temperature drops on average 30 minutes earlier below zero and remains on average 40 minutes longer below zero than the reference pavement. However, one shall note that the surface is so soft that tyres rolling on it will break any ice or packed snow that may form on the surface so, in fact, in black ice conditions, the PERS seemed to be much less slippery than conventional dense asphalt. The winter behaviour is different from that of conventional impervious pavements but not very different from that of porous asphalt, and can be handled with an adapted winter maintenance strategy, as described in the project.
Generally, the durability of the full-scale test tracks was and is still an issue. In spite of the laboratory results, some ravelling appeared on the test tracks, except on the Belgian test track, which performed excellent for this criterion. It is not clear yet why the Belgian test section showed a good ravelling resistance while the Danish test section did not do so, as the mixes used were almost the same. It might have been due to quite different traffic volumes and conditions.
Delamination from the sub-layer was another issue: on all test tracks where the PERS layer was glued on a bituminous sub-layer, the PERS layer delaminated after 9 to 14 months after construction. The Polish test track disintegrated already after a week, but this was probably due to a construction error. It is likely that the action of heavy vehicles plays an important role in the delamination of PERS glued on asphalt. PERS on an asphalt sub-layer on a road without heavy vehicles could probably survive much longer. PERS on a semi-flexible sub-layer (as on the Swedish test tracks in Sjögestad) and glued with epoxy performs much better in this respect and is the only one that still exists when this report is drafted (January 2016). The final tests in the laboratory on the large wheel tracking device, simulating the action of heavy vehicles tyres combined with other influences such as thaw-frost cycles, brine and UV, confirmed the findings on the full-scale test tracks: the binding with epoxy on a semi-flexible or concrete test track is much more durable.
An extensive study and measurements during the construction of the Belgian test track prove that there is no problem with toxic emissions during (or after) the curing of the polyurethane. The smoke produced during a car fire on PERS has been analysed and is not more toxic than from other common materials which are burning.
A surprising result is that the fire risk is not increased with PERS; rather the contrary. Fuel spills are much less dangerous on PERS than on a conventional impervious pavement, such as dense asphalt or SMA. Instead of a few seconds, it takes several minutes before a car above a fuel spill starts to burn on PERS compared to dense concrete.
Provided an equally even surface is obtained, PERS reduces the amplitude of ground borne vibrations typical for heavy vehicles driving on an uneven surface, but only moderately.
The production of fine dust by the action of studded tyres on a PERS pavement is 95 % less than on a conventional pavement.
The influence of PERS on waste management is positive according to a study conducted within this project and there appears to be no problem with surface or ground water pollution by the leaching of toxic products, as measurements have shown.
PERS in its present form is less environmentally sustainable than a conventional pavement, but when one takes into account the noise reducing effect and also considers alternatives for the high noise reduction obtained with PERS (e.g. noise barriers), then PERS appears to be the most sustainable solution. The main implementation of PERS is intended to be as an alternative to noise barriers or where no other noise-reducing options are feasible; thus it will normally be a sustainable option too.
An extensive Cost-Benefit Analysis has shown that PERS can be the best socio-economic choice, namely in situations with a high population of noise exposed people, which results in a high noise cost. In a lot of cases it is impossible to build noise barriers and then PERS is the only feasible option if one wants to reduce the traffic noise.

Project Context and Objectives:
The PERSUADE project aimed at developing a durable, cost-effective Poroelastic Road Surface (PERS) using used tyres, which would at the same time benefit the environment by not only significantly contributing to abating traffic noise and vibrations but also helping to solve the problem of over three million tons of used tyres having to be dumped or burned every year in the twenty-eight Member States with the consequence on ground and air pollution. It was anticipated that an advanced optimization of the poroelastic rubber compound could also lead to a decrease of rolling resistance for the vehicles rolling on such a pavement and, as a result, could contribute to reducing CO2 and other emissions.
Indeed, end-of-life tyres are a disposal problem regarding the large volumes produced every year. Tyre shreds are primarily produced to reduce the transportation volumes of end-of-life tyres after collection. Within the European Union, there is a ban on land filling tyre material in order to reduce the total land filling volumes and to encourage recycling measures [Directives 1999/31/EC & 2003/33/EC]. Until shortly before the start of the project in 2009, the main disposal option has been energy recovery in industrial processes. However, legislation acts has been taken around that moment in the European Union to encourage recycling and recovery of end-of-life-tyres and re-use of tyre materials in construction works is listed as one disposal option. Tyre shreds possess interesting technical properties that could be beneficially used in civil engineering applications. Some characteristic properties of tyre shred materials are the low density, high elasticity, low stiffness, high drainage capacity, high thermal insulation capacity and intrinsic wear resistance. These properties open up possibilities for utilisation of the material in an innovative manner.
In this project, one took advantage of the experience which pre-existed in Sweden and in Japan. The former country is represented in the Consortium while the latter was represented in an External Reference Group (ERG). Five countries hosted the experimental sites and have applied different variants of mixes and construction methods: in situ produced mix, slabs (1 m x 0.5 m) produced in the factory and which were consequently glued to the sub layer and small prefabricated slabs which were glued in the factory on cement concrete setts which were consequently laid in the traditional manner. A new and innovative Dutch method, namely the “Rollpave” system, was originally also foreseen to be tested with PERS, but was in the course of the project abandoned, as partner Duravermeer, the owner of the concept, did not want to proceed with the development of the concept due to some disadvantages. The countries in which test sections were built are: Belgium, Slovenia, Denmark, Sweden and Poland. Their geographical spread allowed for the influences of climate and traffic differences to be tested. Regarding traffic, some countries tested PERS in a city-like street with slow, light traffic while others will test it on a secondary road with faster and heavier traffic.

Project Results:
According to earlier research projects in Sweden and Japan, the poroelastic road surface (PERS), a pavement made of rubber granules and stone or sand aggregates, bound with a flexible polymer like polyurethane, is capable of reducing tyre/road noise by substantially more than any other type of surfacing. Earlier attempts to develop this concept into a low noise pavement ready for application in full scale and meeting reasonable requirements regarding safety and durability have failed. Reasons for the failures have been different in various experiments, but durability and/or safety have been the major problems. Never-theless, this project team had the idea that working together in a large European cooperation project and utilizing all kinds of pavement engineering specialists would be the way forward. In 2009, the European Commission accepted a proposal named PERSUADE (PoroElastic Road SUrface for Avoiding Damage to the Environment) to make a serious attempt to turn the concept and earlier practical experience into an implementable noise reducing product.
The PERSUADE project has the intention to develop the PERS concept into a durable, highly noise reducing, safe and cost effective pavement which can provide exceptional traffic noise reduction under certain traffic and road conditions and thus provide European road and environmental authorities with a new tool for managing hotspots in noise exposed areas.
The following pages summarizes the project work and results with focus on technical issues.
Mix development
The first step to achieve the goal is to develop an appropriate PERS mix formulation. The properties that had been defined as targets for the mix optimization included:
• durability and stiffness of the material in the field
• frictional properties during the service life and under various climatic conditions
• noise absorption and noise generating properties in the interaction with tyres
• drainage properties
• rolling resistance
• emissions of particulate matter and hazardous materials to the environment
To this end, a number of laboratory experiments were conducted to explore the mentioned properties of several interesting mixes. Regarding the stiffness, the results indicated that all the PERS mixes, regardless of the different compositions of each type, behaved similar in that the first cycle always was partly non-elastic (higher deformation incurred) after which, during the next consecutive cycles, the samples were stable against higher hysteresis losses. Under real traffic scenarios, one can expect that the higher the rubber content, the less the plastic (or non-elastic) deformation incurred. Comparison of dynamic moduli was undertaken between the PERS mixtures and a typical asphalt concrete mix showed that the asphalt mix had about 200 up to about 1000 times higher dynamic modulus than two types of PERS which were later tested on the road. This showed that PERS mixes were much softer in nature, which could provide a potential for less noise produced in contact with rolling tyres.
As for polishing resistance and friction after some time of polishing, it was concluded that the evolution of friction with polishing on PERS could be studied without major difficulties. The main conclusions of the laboratory experiments are:
• The maximum friction value increases with the proportion of hard aggregates in the mix.
• The PERS mixes developed and intended for later field testing show first an increase and then a quite constant value of friction as they are polished. At the initial states, the maximum friction of PERS is smaller than maximum values on common French road surfaces, but the friction of some PERS mixes becomes close to DAC 0/10 or VTAC 0/10 surfaces after 50 000 cycles of polishing.
These results are encouraging for implementing a safe and durable PERS surface in terms of skid resistance. Next, research should focus on the increase of the initial and maximum friction value of PERS surfaces, possibly by adding sand to the polyurethane (PU) binder.
The optimization of the poroelastic mix to improve its ravelling resistance, on the other hand, turned out to be a real challenge and required the study of a lot of various mixes in the PERSUADE project. Most of the testing utilized the Aachener Ravelling Tester (ARTe). To evaluate the perfor¬mance of the mixes in a large scale production; some of them were sampled from production in field tests, to come one step closer to reality.
The results show significant differences in terms of ravelling resistance between the tested mixes, sampled from different test sections. It is not possible to give a definite explanation why the results from the different test sections differ so much. For example, the results for the mix used in the Belgian test section laid at the end of the project fulfilled the objective for the ravelling resistance at the beginning, as described in the results section. The on-site properties in terms of ravelling resistance have to be studied in future experiments to confirm the results of ARTe testing, since the trials in PERSUADE could not be continued as long time as desired.
Experiments to study the wear under simulated traffic were carried out in the VTI circular road simulator and led to the following conclusions
• The wear and raveling resistance of the best performing PERS material tested is equal to the wear resistance of a high quality SMA with 16 mm maximum chipping size.
• The factors that greatly influence the wear resistance are:
▪ Fraction of aggregate (lower is better)
▪ Air void content (lower is better)
• The type of aggregate has some influence on wear resistance (Jelsa granite performed 20 % better than the Skärlunda granite or Forserum diabase)
• For factory produced material the wear resistance is not improved if the binder content is raised from 9 to 11 %.
A comprehensive risk analysis was carried out concerning the risks involved with the exposure to emissions produced from PERS. It was found that there is an adequate control of the risks associated with production and scrapping of PERS for the workers and the general population. This is the case for both planned operations and the non-recommended operations such as placing hot poured asphalt on remnants of PERS.
An experiment was set up to study the particle emission from the interaction of studded tyres with PERS compared to conventional pavements and it was concluded that:
• A number of PERS variants resulted in 10-20 times lower PM10 production and about 90 times lower number concentration (at 70 km/h) than a standard asphalt (SMA16) with granite.
• The mass size distribution were bimodal with peaks at 2-3 and >8 µm, and in mean size slightly coarser than from the reference asphalt, but indicates that the source is dust from wear of pavement rocks.
• Number size distributions show that ultrafine particles are produced during the test and that these are fewer and smaller than the ultrafine particles formed when wearing the reference asphalt. The number concentration, though, is very low and comparable to normal background values.
An extensive test was carried out to assess the durability under “harsh” real life-like conditions with multiple influences, comprising brine, water, freeze-thaw cycles and UV exposure, combined with the interaction of the tyres mounted on a large “wheel-tracking device”. This showed that the tested PERS material had a durability comparable to or better than that of a conventional thin layer asphalt. However, the wheel tracking test also showed that the samples submerged in water also resulted in a delamination of the PERS mix from the asphalt concrete mix.
Structural design
A second important issue was the structural design of roads with a PERS wearing course. A wide range of tests was performed related to this topic, both in–situ and in the laboratory. Due to the typical low stiffness of PERS mixtures compared to the ordinary asphalt mixtures, not all planned tests can be expected to give similar results. The results were used for the calculation of the influence of a PERS wearing course on the durability of the road.
Furthermore, the PERSUADE consortium has provided a procedure for the pavement design with a PERS wearing course and to estimate the influence of such a layer on the pavement life. The developed procedure allows for the proper design of the pavement. The following critical phenomena were taken into consideration:
• the influence of the PERS wearing course stiffness on pavement design life, and
• the influence of bonding between PERS and the base course on the pavement design life
All (planned and actually used) full-scale test section locations were tested by means of Falling Weight Deflectometers (FWD). The results showed that in most cases the calculated mean deflections at the test section before application of PERS were below 0.5 mm. According to the accepted methodology, it can be concluded that the bearing capacity should be good enough for at least medium traffic category, up to 7.3 millions of standard 100 KN axles in a 20 years design period. This means that these sections did not need to be reinforced before laying a PERS course. In only one case, a planned localization in Denmark, the results of measurements indicated that the thickness of new asphalt layers (upgrade) should be 12-16 cm depending on the location. Finally, but for other reasons, that test site was moved to another location.
In Belgium, FWD measurements were performed before and after laying a PERS wearing course. In the “after” measurement, the calculated deflection increased approximately ten times compared to “before” (laying of PERS). According to the nomograph used in Poland the thickness of new asphalt layers (upgrade) should be over 24 cm on the area of the test section. The results show that due to the elasticity of PERS, the calculation of the bearing capacity on the basis of FWD measurements is inappropriate and in reality the road condition is better than estimated.
FWD measurements should be performed on the existing pavement without PERS layer and the calculated deflection should be below 0.5 mm. This is the proof that the influence of PERS course will be negligible in order to avoid problems with pavement durability.
The so-called Leutner test used to assess the shear strength, yielded a value of 0.4 MPa, which is insufficient to pass Polish requirements, which means that an improved solution should be found. It should be noted that the deflection observed at the maximum force (appr. 5 mm) is 3-7 times higher than in case of standard asphalt mixtures. Moreover, a total delamination was not observed in the Leutner test.
The hollow cylinder test performed on the hollow cylinder apparatus (HCA) enables torsional simple shear loading, which includes rotation of principal stresses. A hollow cylinder of pavement material can thus be tested by applying a combination of axial and torque loading, subjecting an element of material in the specimen wall to a realistic stress path. Such test reflects to a relatively high degree effects of real traffic on materials used in a road pavement.
PERS material is typical elasto-plastic material, and such a behaviour can indeed be observed during cyclic loading. One can see recoverable and irrecoverable strain during loading. Besides the range of strain, stiffness and damping are also very important para-meters for deformation behaviour characterization. So-called hysteresis loops develop during loading. Based on this curve, shear modulus and damping of material were evaluated. Initial perfect elastic behaviour changed in later loops to plastic behaviour.
The equivalent shear modulus of PERS material and hysteretic damping ratio were evaluated for various stress conditions. Finally, it was observed that stiffness is decreasing almost linearly with an increase of strain, while on the other hand damping is increasing.
Test sections with PERS
The above findings regarding mix formulation and structural design were applied in a large campaign of building of test tracks. First six small-scale test tracks were built, having an area between 10 m² up to 60 m²). They have been built on locations with almost no, little or moderate traffic in four partner countries and using three different laying techniques:
• in situ mix
• prefabricated slabs of 0.5 m x 1 m glued on the sub-layer
• small prefabricated slabs glued in the factory on cement concrete blocks
The pilot test tracks were intended for monitoring a limited number of important parameters (such as skid resistance, winter behaviour, durability) and allowed contractors to practice with the novel material and application techniques.
The following “full-scale test sections” were constructed:
• The first Danish test section (PERS-DK1) was built in Kalvehave, covering 75 m x 3.5 m and 30 mm thick, using the in situ laying technique with a prior tested “Arnakke” mix, in August 2013.
• The second Danish test section (PERS-DK2) was built on the same location and with the same dimensions in June 2014, following a failure due to excessive ravelling of the first test section. For the second test section an improved mix was used (addition of cellulose fibres as anti-dripping agent).
• The first Swedish test section (PERS-SE1-HET1) was built in Linköping, only in the right wheel section with dimensions 25 m x 1 m and 30 mm thick. It was built in November 2013 on a street in Linköping. The partner HET produced prefabricated slabs (0.5 m x 1 m) which were glued to a new asphalt sub-layer with epoxy, but the laying conditions (beginning of winter) were far from ideal, although the construction was carried out under a tent. The mix used was the so-called HET mix, with a higher (about the double) amount of rubber.
• The second and third Swedish test sections were constructed on a main road in Sjögestad in Linköping in August – September 2014. PERS was applied in 1 m wide strips in both wheel tracks and surrounded by porous asphalt, in an attempt to protect the PERS against damage from snow ploughs and at the same time to provide a proper water drainage. For one (PERS-SE2-HET2), 30 m long test section prefabricated HET slabs were used, and for the other (PERS-SE2-VTI) 24 m long test section an in-situ laid PERS was used. The thickness of both test sections is designed to be 30 mm. A special feature of this test site is that a dedicated “semi-flexible” sub-layer was constructed instead of the old asphalt, whereas on the other full-scale test sections (except the Slovenian one) one had new or old (milled) asphalt concrete as sub-layer. The semi-flexible layer is a kind of hybrid of an asphalt and a cement concrete layer, and is used in locations were exceptional strength is required, such as at bus stops.
• The Belgian test section (PERS-BE1) with dimensions 40 m x 3.5 m and a thickness of 45 mm was built on a rural road in Herzele in September 2014 by means of the in situ technique. The same mix was used as the second Danish test section. The sub-layer was a milled dense asphalt concrete in rather poor condition.
• The Polish test section (PERS-PL1) was also built in September 2014 on a road in Kryspinów near Kraków and the dimensions were 65 m x 3 m with a PERS thickness of 25 mm. The mix was similar to the mix used in Denmark and Belgium, but with local Polish aggregates. The in situ technique was used. The PERS was laid on a new dense asphalt concrete as sub-layer.
• The full-scale Slovenian test section (PERS-SI2-HET2) was constructed adjacent to the earlier built pilot test section on a street in Nova Gorica in December 2014. The main part with a length of 18 m x 3 m was constructed with the same technique as the pilot test section: i.e. prefabricated PERS tiles glued on cement concrete blocks. A section was constructed with a length of 2 m with the tiles in situ glued on a cement concrete sub-layer (hence without cement concrete blocks).
As damages - unfortunately - did occur on the full-scale test sections, the team took advantage of this to practice with techniques to repair loosened patches and assessed and reported about these repairs.
Monitoring of the test sections
For each of the test sections, immediately after its construction an intensive monitoring programme started, in an attempt to extract as much knowledge from it as possible. A variety of parameters were monitored, which are treated one by one in the following:
• Noise
Noise measurements, both of the SPB and the CPX type, have been performed on the test pavements. Different reference pavements are used around Europe. In this project four references have been included. If the Swedish SMA 16 is used as a reference, the noise reduction (for SPB results) for the PERS-SE2-HET2 and PERS-SE2-VTI, with the smallest aggregate size and the largest proportion of rubber aggregates in the test program, was 12 and 11 dB when the pavements were new. For a studded tyre, even 16 dB was measured. The other four PERS pavements (with larger aggregate size and less rubber aggregates) included in the SPB measurements had a noise reduction of 8 to 10 dB during their first month. If the Danish SMA 11 is used as a reference, the noise reduction for the PERS-SE2-HET2 and PERS-SE2-VTI was 10 and 9 dB when the pavements were new. The other four PERS pavements included in the SPB measurements had a noise reduction of 6 to 8 dB during their first months. The general picture is that there is a tendency of around 1 dB higher noise reductions at 80 km/h than at 50 km/h.
These results are generally confirmed by the CPX measurements. The CPX measurements also included the PERS-PL1 as well as the PERS-SE1-HET1 pavements. The result was that they had the same high noise reduction as the PERS-SE2-HET2 and PERS-SE2-VTI pavements. The measurements also indicated that the PERS pavements also provide a noise reduction for heavy vehicles. But for this vehicle category the noise reduction (measured with the SPB method) is 2-4 dB less than the noise reduction for passenger cars.
• Sound absorption measurements
The acoustical absorption was measured on circular specimen drilled from test slabs of the eight PERS pavements. More than one drill core representing every PERS pavement have been used for the absorption measurements. The drill cores from the factory-produced HET plates are similar to the paved surfaces, but some minor differences between the compaction of the in-situ produced surfaces and the test slabs are expected.
Measurements are carried out according to ISO 10534-2 in an impedance tube. The acoustical absorption is measured on two to six cores drilled from each test slab.
The general impression is that the PERS-SI2-HET2 and PERS-SE2-HET2 surfaces have similar sound absorption frequency spectra. Both have 4 mm maximum aggregate size and are produced in a factory. The PERS-DK2 and PERS-BE1 have similar spectra; likewise the PERS-PL1 spectra are similar to each other with maximum peak frequency outside the measured frequency range. The PERS-SE2-VTI deviates from the other surfaces by having a high absorption in a wide frequency range, with the peak in the frequency range of 1600 Hz.
The standard deviations unveil that there is a large spread in both peak frequency and absorption coefficients. This is particularly the case for the PERS-BE1 and PERS-SE2-VTI, which indicates that the test slabs from these surfaces are inhomogeneous.
The design goal for normal porous asphalt is to get the maximum absorption between 600 and 1000 Hz, which has been achieved for five of the test pavements according to these absorption measurements.
• Rolling resistance
Rolling resistance coefficients have been measured on five of the PERS test pavements. Asphalt concrete or SMA pavements used on a road nearby the test sections were used as reference pavements. The five PERS pavements had higher rolling resistance coefficients than the selected older reference pavements.
The two PERS-DK1 and PERS-DK2 constructed on-site with paving machines had the lowest increase of rolling resistance (7 and 5 %) whereas the PERS-SE2-VTI that was manually laid had an increase of 30 %. The PERS-SE2-VTI got an uneven surface structure due to being manually laid and that might be the reason for the high rolling resistance coefficients.
The PERS-SI-HET2 and PERS-SE2-HET2, constructed with small and large HET slabs, had a higher increase of rolling resistance than the two on-site-constructed PERS-DK1 and PERS-DK2 pavements.
• Various structural and geometrical features
A thin and plane section analysis was carried out, although it is not related to the acoustical performance. However, the results are of great importance regarding the durability of the surfaces. The thin and plane analysis illustrates that the stone and rubber aggregates generally are covered with polyurethane, which is a positive factor for the durability of the pavements. The analysis also shows that there is a big discrepancy between the different samples despite some being taken or constructed from the same recipe.
Texture was measured on the test sections with a laser profilometer and the texture spectra of the surfaces were calculated according to ISO/TS 13473-4 and the MPD values according to ISO 13473-1. Due to the short lengths that the profilometer was able to measure the spectra only contain information about the macrotexture. The spectra of the five surfaces are generally similar, with PERS-BE1 having a slightly lower level at the shorter wavelengths.
Permeability/draining capacity was measured with the Becker tube and reported. They varied a lot from location to location and were generally lower (higher outflow times) than expected for such high air void contents.
• Mechanical impedance
The mechanical impedance of PERS samples taken from the eight full-scale test sections has been tested in the IFSTTAR laboratory. It was found that the mechanical impedance of PERS was very different to that of asphalt pavements, and the softest PERS samples were just a little higher than the mechanical impedance of tyres.
• Monitoring winter performance.
It is very important to know how these new poroelastic pavement types behave when exposed to typical European winter weather conditions and how they react to special maintenance during winter.
An extensive thermal behaviour monitoring has been carried out during two additional con-secu¬tive winter periods (the last winters in the project period) on the Sterrebeek (BE), Herzele (BE) and Kalvehave (DK) PERS test sections. This monitoring to a large extent confirmed the earlier findings that that PERS has very typical winter behaviour, different than the one of the dense asphalt concrete used as reference on all three sites. The Swedish sites were also frequently monitored (visually), but not quantitatively.
Under calm weather situation with a clear sky, the surface temperature amplitude was several degrees higher (how many depends on the radiation conditions) on the PERS as compared to the reference pavement (i.e. PERS has lower heat capacity and heat conduction). On the Belgian test sections the temperatures registered by the temperature sensors 50 mm under the surface in the reference pavement and in the PERS showed that the amplitude below the PERS is smaller, meaning that it more or less behaves like an insulating layer. These considerations indicate that the PERS surface layer benefits less from the ground temperature (as compared to the reference).
Several cooling and warming events have been analysed during both observation periods. They indicate that the PERS pavement anticipates negative temperatures (around 30 minutes on average before the reference pavement) and needs more time (around 40 minutes on average than the reference pavement) to recover positive temperatures. However the advances and delays could be quite different from one period to the other; this mainly depends on the prevailing atmospheric situation and particularly the sky coverage. Periods characterised by sudden drops and rise of the surface temperature (clear sky) usually cause larger advance and delays.
A comprehensive report about the winter monitoring has been produced, describing the special winter behaviour of this pavement type and prescribing how to deal with it.
• Visual inspection
Regular visual inspections were carried out on all the test sections. Based on the visual inspections, it must generally be concluded that in relation to durability none of the test sections met reasonable requirements. However, on roads with essentially only light vehicles and if snow ploughing can be controlled in some way; some of the test sections reported here might have provided reasonable durability.
• Friction
The wet friction was measured using different measurement methods; the Pendulum, the Portable Friction Tester (PFT) and national friction measurement devices. Results from the three different measurement methods cannot be compared. For three of the PERS pavements, special action, such as applying sand and/or grinding have been used successfully to improve the friction immediately after the laying of the pavements. The overall conclusion of these friction measurements is that seven out of eight PERS pavements fulfil the national requirements for friction (after the initial treatment needed for three of the pavements).
Environmental impact
The “environmental impact” of PERS was extensively assessed in the PERSUADE project. These are treated here for each feature separately.
• Fire risk
A comprehensive study has been made of the fire risk, both assessing the rapidity of the spreading of the fire and the risk of the emission of toxic gases. Small-scale laboratory, medium-scale and full-scale tests (burning of a car on a PERS) have been carried out. It turns out that no special risk has been identified regarding toxicity of the smoke during a fire. And – surprisingly - in the case of a vehicle fire combined with a fuel spill, PERS turns out to be much safer than a conventional impervious pavement. Whereas it takes seconds to get high flames on a dense asphalt, it is a matter of several minutes before a car on PERS, in both cases with a 20 l fuel spill under it, catches fire.
• Drum tests of noise emission and rolling resistance
Samples of PERS- materials produced by partner HET were glued on the drum of TUG's drum facility in order to perform rolling resistance and noise tests. The results showed that PERS is durable and passenger car tyres did not cause extensive erosion of the surface, which complies with the findings on the ARTe (ravelling testing). The noise reduction was considerable and rolling resistance was very low.
• Vibration damping with PERS
The potential of poroelastic road surfaces (PERS) as a vibration abatement measure has been assessed and it has some influence on ground-borne vibrations, but rather moderate.
• Impact on CO2 balance
Measurements of rolling resistance (and tyre/road noise) were performed with a variety of tyres and pavements on drum facilities at the laboratory of the Technical University of Gdansk (TUG) in Poland and with a few reference tyres on road test sections located in Denmark, Sweden and Slovenia..
It appears that the rolling resistance is somewhat higher on PERS than on DAC pavements but this may be mainly since the trial sections are more uneven than conventional pavements, and this is possible to improve on in future trials. For truck tyres the rolling resistance appears to be 30-40 % higher than on conventional asphalt pavements.
The impact of PERS on the global fuel consumption and CO2 emission will, anyway, be marginal, as even when it is fully operational it will cover only a small fraction of the road network. It is intended for abating noise mainly on so-called hot spots.
• Waste management (recycling, availability of material, final waste disposal)
A study was conducted on this subject and, considering all factors, it can be concluded that the material balance for the PERS system is a positive one: it prevents the first end-of-lifetime options (such as landfill or incineration) for post-consumer rubber tyres, and prevents the production of new mineral aggregates from quarries to be incorporated in the original asphalt mixtures for road pavements. On the negative side, one has to account for the consumption of PU binder, compared to the use of bitumen in regular asphalt pavement mixtures.
At the second end-of-life stage (of the PERS itself), suitable options exist for a sustainable treatment, such as recycling in new asphalt mixtures, or incineration with energy recovery.
With its substantial noise reduction, the PERS system prevents the need for production and use of other materials that would be used to build noise barriers, as well as the energy and labour necessary to install these along the roadsides, in order to get the same noise-attenuating results.
• Pollution of surface and ground water
A sample of PERS was crushed and in granulate form subjected to a column leaching test (CEN/TS 14405) and an equilibrium column leaching test (NT TR 576) in order to determine the leaching of inorganic substances and organic conta¬minants, respectively.
The leaching of inorganic substances (chloride, fluoride, sulphate, As, Ba, Cd, Cr, Cu, Hg, Mo, Na, Ni, Pb, Sb, Se and Zn) at L/S 2 l/kg was significantly lower than leaching reported for recycled crushed asphalt, except for Mn, which showed five times higher release from PERS compared to the 90th percentile reported for recycled crushed asphalt. The release of Mn from PERS at L/S = 2 l/kg exceeds the current Danish leaching limit value for recycling and using certain residues for e.g. road construction by a factor of 3 to 4. However, considering that PERS will be applied in only 3.5 cm thick layers, this may not constitute a real problem.
Measurable solution concentration of Sn (which was used as an indicator instead of measurement of organotin compounds) was found only in the first eluate (at L/S 0.1 l/kg) from the column leaching test (0.38 µg/l corresponding to a release of 0.000038 mg/kg). The calculated released amount of Sn at L/S 2 l/kg and L/S 10 l/kg was <0.00023 mg/kg and <0.0010 mg/kg, respectively. Even though the concentration of Sn appears to decrease quickly with L/S (or time), this indicates that the release of Sn/organotin may be an issue that should be looked into in more detail and/or taken into account if it is decided to apply PERS in a major scale. Again, from an environmental perspective it is a positive factor that the PERS is applied only in relatively thin layers.
The release of petroleum hydrocarbons from PERS was significantly higher in all carbon intervals than results reported earlier for recycled crushed asphalt. The highest release was measured in the C25-C35 range which is normally referred to as the “oil-range”. In addition, it should be noted that the laboratory reports a “presence of hydrocarbons of unknown origin”. The issue was not investigated further in this project. A comparison with recently performed risk assessments on the application of crushed asphalt in different road construction scenarios indicates that, as far as total petroleum hydrocarbons (TPH) are concerned, the PERS material would most likely comply with criteria considered adequate to protect groundwater quality directly below roads of 4 to 8 m width (probably also broader), when applied in layers of 30 cm (as compared to the actual layer thickness of 3.5 cm) and when the percolation of precipitation through the material is restricted to 70 to 250 mm per year with an assumed thickness of the unsaturated zone of 1 m.
The leaching of polycyclic aromatic hydrocarbons (PAH), including five specific PAHs prioritised by Danish EPA, from PERS was comparable to or lower than earlier reported results for recycled crushed asphalt.
• Life Cycle Assessment
In order to assess the relative environmental sustainability of PERS, a comparative life cycle assessment (LCA) has been performed on PERS as compared to conventional pavements, i.e. SMA. In this first ever LCA on PERS, an attributional (average) approach has been used including cut-off modelling for end-of-life. In total, three different life cycle impact assessment methods have been used in the assessment, covering both midpoint (impact) and endpoint (damage) methodology.
Results show that for almost all impact or damage categories included, the potential impact or damage of the PERS pavement is higher than that of the SMA pavement. When the difference in the noise emission is taken into account, by combining the conventional pavement with a 3 m high noise barrier, the results are turned upside down and then the PERS pavement shows best environmental performance regarding both a one lane road (one lane in each direction) and a two lane road. However, this result is based on the assumption of equal lifetime, i.e. 15 years. The lifetime of the most promising PERS pavements, for the full-scale test sections included in PERSUADE, is estimated to be maximum 2-3 years. Therefore, an analysis of the breakeven for the PERS wearing course life time was performed by use of single score damage indicators. The outcome indicates breakeven life times at 3.8 – 9.2 years depending on the number of lanes and the height of the noise barrier (3 m or 4 m). If the SMA pavement is substituted by a proxy for porous asphalt with a life time of 7.5 years, the lowest PERS pavement breakeven lifetime is obtained for a one lane road and in this case it amounts to 3.4 years.
The results of this LCA study indicate that PERS based on polyurethane binder systems are less environmentally sustainable than conventional pavement types such as SMA based on bitumen binder systems. But when the noise reducing effect of PERS is taken into account, by introducing noise barriers for the conventional pavement types, the PERS pavement apparently becomes the most environmentally sustainable. However, that is based on the precondition of equal lifetime among the two compared pavements or at least a lifetime of around 4 years or higher for the PERS pavement.
Cost-Benefit Analysis
It is important that PERS is interesting from an economical point of view and to assess this a comprehensive Cost-Benefit Analysis has been carried out. Two analyses have been carried out: a preliminary one to establish the model but with “preliminary” figures, and the end of the project figures obtained from the test sections have been put in the model.
The conclusion from the cost-benefit and the sensitivity analyses is that, given the assumption made for the various parameters, PERS can be a socio-economically more efficient choice compared with the reference surface, but only in cases where external noise costs are high and where many people per km are affected by noise.
In the initial analysis, the possible influence of using PERS has been ignored for three parameters: recycling of tyres, risk of accidents and air pollution. These three elements pull in opposite directions.
Drawing on studies on recycling of tyres, it appears that it is no problem to find other uses of rubber from worn-out tyres, and since reuse of tyres has no value in itself, this element will not change the results of the analysis.
In terms of accidents, studies have shown that tyre/road friction is still acceptable and is not affected in a way expected to cause more accidents. Therefore, we do not expect that the PERS pavement will affect safety.
The rolling resistance of PERS may increase significantly for the somewhat uneven test sections produced so far. However, due to the marginal road distances that may be covered with PERS in a future implementation phase, there is no reason to believe that fuel consumption and thereby CO2 emissions and air pollution would change significantly in an overall picture.

Potential Impact:
Potential impact
The PERSUADE project has demonstrated that the use of poroelastic road surfaces (PERS) has a substantial potential for improving, or in other ways influencing, some of the environmental policies of the European Union (EU), provided that the expected durability can be demonstrated in post-PERSUADE studies and some remaining issues can be resolved.
The effects of implementing PERS on European roads on the EU environmental policy issues are explored and discussed in this report. The results of the project, which are relevant for the various environmental fields of interest of the Commission, are also addressed. Examples of such issues are the following:
• Noise reduction of PERS is relevant for the Action plans as foreseen in the European Noise Directive (2002/49/EC)
• The noise properties of PERS is relevant to the new common noise assessment method (Commission Directive (EU) 2015/996); in particular the effect of the road surface
• CO2 emission related to the production, construction and use of PERS is of particular interest for the commitments of the EU to reducing emission of greenhouse gas
• Reuse of worn tyres for the production of PERS is of interest for the objectives stated in the Directive 2000/53/EC on end-of-life vehicles.
The conclusions below relate to when PERS is imple¬mentable; i.e. where durability has been improved to acceptable levels but with other features complying with the effects achieved in this project. The following conclusions related to EU policies are drawn from this deliverable:
There is no other noise-reducing measure with such exceptional efficiency as PERS, except very tall noise barriers or covered roads (tunnels), and the noise reduction is achieved without the disadvantages of noise barriers.
The use of PERS is very positive in relation to the intentions of the landfill, the recycling and the ELV directives, since it creates a “second life” for rubber from end-of-use tyres.
Although the EU 2020 strategy to reduce overall greenhouse gas emissions by 20 % compared to 1990 levels does not specifically deal with road surface rolling resistance, it must be acknowledged that PERS in the field tests appeared to have an increased rolling resistance compared to common pavement types, and thus more fuel would be consumed and more CO2 and other exhaust gases would be emitted. However, since PERS is unlikely to be used other than in “hot spots” (road sections with exceptional noise exposure) the effects will be at most marginal.
Since the EU has the policy to reduce the fleet of fossil fuel depending vehicles with electric vehicles, it is interesting to note that the noise-reducing effect of PERS will increase when noise from the power units of vehicles (“propulsion noise”) will gradually decrease with the shift towards electric vehicles; especially for heavy vehicles in urban areas. Already today, a significant part of the public bus network in urban areas is operated by electric or hybrid busses.
PERS offers an exceptional noise reduction, which could be very useful when it comes to satisfy the noise-related directives, such as the environmental noise directive (END) and the new directive on uniform calculation of road traffic noise (“CNOSSOS”).
In cases where noise barriers are ineffective, which often occurs in urban areas, PERS may be the only practical measure to reduce noise propagating from the street to the environ¬ment, and it would work without causing the often very undesirable obstruction of view from homes near the street or road, and with similar effects in all directions from the road.

Optionally, there is also a potential effect if PERS is used to allow closer distances between roads and residential areas without creating negative environmental impacts. The potentially up to 10 dB by which PERS can reduce noise may release large areas for development, which otherwise might have too high noise exposure to be suitable for residential exploitation.
An unexpected positive effect of PERS, related to the tunnel safety policies in Europe, is the dramatically reduced fire danger on PERS compared to dense concrete or asphalt surfaces in cases of spilled fuel from vehicle collisions that is starting to burn.
Dissemination activities
A variety of dissemination activities have been undertaken:
* by means of the regularly updated project website
* the mid-term and final seminar
* 21 international conference papers
* 4 national conference papers
* 5 international conference presentations (without paper)
* 2 national conference presentations (without paper)
* 3 articles in scientific journals
* 8 articles in popular national or regional journals
* 13 media activities (i.e. radio and television covering)
* a presentation for the Working Group "Noise" of the UNECE in Geneva
Exploitation of the results
* Where is PERS most suitable for implementation?
The experience of PERSUADE is that the following should be observed when implementa¬tion of PERS on actual roads or streets is considered:
• Use PERS on roads where traffic is highly dominated by light vehicles. Due to durability issues, do not use PERS where traffic of heavy vehicles exceeds 10 % of the AADT; preferably not even above 5 %.
• As for all kinds of porous pavements, use PERS only when proper drainage to the roadside can be provided. Proper drainage can be provided in different ways; for example by extending PERS also over the road shoulder, by using special kerbstones with drains, by gullies or other pavement drains in the PERS, or by using perforated pipes with outlet to the roadside (see above). In all cases, road crosslope should be extra high; at least 3 %.
• Use PERS where high-performance porous asphalt (normally a double-layer) would be an interesting noise-reducing measure but where clogging by dirt is expected to be a big problem, since PERS is to a large extent “self-cleaning”.
• Since clogging is a huge problem for porous asphalt pavements on low-speed roads and streets in urban areas, PERS could be an interesting option; in particular when no other measures to reduce noise emission or propagation are practical or efficient.
• Use PERS where noise from studded tyres is a big problem, since it eliminates virtually all tyre/road noise in such cases, but make sure that snow-ploughing is made in a way which does not damage PERS.
• PERS has not been tried in locations with stop-and-go traffic and at intersections. Durability may be extra critical in such locations, as it is for all porous pavements. Implementation at such locations is not recommended until testing has confirmed that it is durable also there.
* Readiness for implementation
Implementation of PERS in full scale on European roads is not yet possible. Since the lifetime of PERS is not yet known and is only estimated, and since construction is critical and needs experience of the contractor, before implementing the PERS in full scale it is recommended to do the following:
• Use the experience collected in this project by reading instructions in the final report and other documents produced in PERSUADE
• Produce a trial mix of PERS or use prefabricated PERS panels
• Make a pilot construction on a short test site on a trafficked road (may be only one wheel track), to get experience (if such experience does not already exist)
• Based on the experience of the above, make a full-scale demonstration track paved with PERS and follow-up with observations and measurements over its lifetime.
* Follow-up work to make PERS fully implementable
In order to make PERS implementable without the preparations mentioned above, it is recommended to do the following work:
• Follow-up the existing PERS field applications in PERSUADE to determine the lifetime and long-term properties
• Using the latest experience, produce a state-of-the-art PERS and monitor its perfor-mance for a considerable time
• Look for more sustainable binders
• Use snow ploughs with less aggressive blades; also consider using GPS data to tailor snow plough settings due to the pavement; implying that on PERS, the blades would be slightly lifted. This would not result in less efficient snow removal, since the flexibility of PERS means that snow does not stick to the PERS surface. There are already existing ploughs with rubber blades, which showed promising properties in part of the project.
Part of this may be made in new or existing national research projects after PERSUADE is finished. It may be mentioned that at least one follow-up project, in the City of Antwerp, is already ongoing, and more results might follow.
In the Netherlands an R&D programme named “Ultra-silent pavement” is already using PERSUADE experience in several innovative applications.
A project proposal "NEREiDE" has been submitted in answer to a LIFE project call in 2015, comprising some further testing on the coherence and adhesion of PERS, the construction of a full scale test section in Belgium and a 500 m test section in Tuscany, Italy

List of Websites:
www.persuadeproject.eu
contact:
Dr Luc GOUBERT
Senior researcher
Responsible Surface Characteristics, Markings and Noise Lab
Belgian Road Research Centre
TEL.: [phone]
[e-mail]