1.0 IntroductionThis paper will examine Renzo Piano’s Jean-Marie Tjibaou Cultural Center, located in New Caledonia, aFrench territory in the South Pacific. The building’s ten wooden cases, referencing the traditional Kanak hutsas well as the surrounding vegetation, create the imagery of the building. These cases are also a highlyarticulated environmental system which allows for natural ventilation of the building. In examining thebuilding’s development, as well as the thermodynamic principles used, it is clear that the cases were notcreated out of a desire for a specific environmental system. Rather, the cases and the remainder of thebuilding were adopted to accommodate this natural ventilation system. This integration of imagery andfunction within the cases prevents them from being mere iconography and binds them to the remainder of thebuilding.

1.1 Project and climate descriptionNew Caledonia is an island in the Pacific Ocean about 1600 kmeast of Australia. The Jean-Marie Tjibaou Cultural Center islocated on a thin peninsula near the island’s capital, Noumea.Running east-west, the 1000 foot long circulation spine runsalong the peninsula’s ridge. On the more protected lagoon sideare four Modernist, flat-roofed glass and steel pavilions. On thebay side are the 10 wooden cases with their curved facadestowards the prevalent winds coming across the ocean (seefigure 1-1 and 1-2)

Renzo Piano received the commission for the Jean-MarieTjibaou Cultural Center, a $33.4 million arts and educationcomplex, through an international design competition held in1991. The center was a gift of the French government to theterritory to promote and preserve the native Kanak culture. Thestructure of the cases, which are up to 28 meters tall, is lami-nated iroko wooden beams with steel cross bracing and con-nections. Natural wood, glass, and steel compose the remain-der of the building. Three different programmatic spaces arecontained within the building. The first, most public spacecontains the entry, a theater, exhibition spaces, cafe, and giftshop. The middle section contains offices for visiting scholars,a library, a computer and media room, and more exhibitionspaces. The third part, located at the end of the peninsula, arethe administrative offices and educational facilities.

Figure1-1: Site plan of the facility (Zabalbeascoa: 6)

Figure 1-2: Aerial view of the complex lookingtowards lagoon in background (Zabalbeascoa: 5)

The climate in Noumea is considered to be ‘oceanic tropical’. There are only moderate variations in tem-peratures from a winter minimum of 65 degrees Fahrenheit to a summer maximum of 93 degrees Fahrenheit.The average relative humidity is about 75% RH with average monthly maximums of 90% RH and minimum of60% RH (Banfi 26) In such a warm and humid climate, ventilation is required in order to supply fresh air, forbody cooling, and for cooling of building. Wind speeds within the building desirable for such tasks are 0.53to 3.04 m/s with 0.28 m/s being the minimum (Bansal 138, table 3.3.5-2).

2.0 Thermodynamic principlesThe Jean-Marie Tjibaou Cultural Center was designed with the desire to utilize natural ventilation. Throughoutits development and in its final design, two main principles are used to achieve natural ventilation: stackventilation and ventilation due to wind forces.

2.1 Stack VentilationThe main principle utilized by stack ventilation is con-vection. The warmer air of the interior rises and leavesthe building from a higher outlet. The heavier, cooler airis replaced via a lower inlet as shown in figure 2-1(Bansal 138).

The rate of flow induced by this thermal forces is givenby the equation:

V = 0.117A[h(ti - ta)]1/2

V = ventilation rate (m3/s)A = free area of inlet opening (m2)h = vertical distance between inlet and outlet (m)ti = average indoor air temperature (Celsius)ta = average ambient air temperature (Celsius)(Bansal 138).

The ventilation rate for a given inlet opening, therefore,can be increased by increasing the vertical distancebetween inlet and outlet (h) or through a greater differ-ence between the indoor and outdoor temperature (ti-ta).These relationships can be see in figure 2-2. Given theclimate in New Caledonia, there would be a negligibledifference between the indoor and the outdoor tempera-tures. Therefore, only by increasing the vertical dis-tance between inlet and outlet, could a desired ventila-tion rate be achieved.

Figure 2-1: Principles of stack ventilation (Bansal: 138).

Figure 2-2: Graph of the relationships between the variablesgoverning stack ventilation (Bansal: 139).

2.2 Ventilation due to Wind ForcesVentilation due to wind relies upon the pressure differential created by the incoming force of the wind. Whenwind strikes a building, a region of higher pressure is created across the side of the building facing the wind(windward side) while the pressure on the leeward side, side walls and roof are reduced. This creates apressure gradient in the direction of the incident wind. Air flows through the building from the openings in theregion of higher pressure to openings in regions of lower pressure (Bansal 139).

This relationship can be understood by looking at an isolatedenclosure with openings on the windward and leeward sides only:

Q = KAV

Q = rate of air flow (m3/h)A = area of smaller opening (m2)V = outdoor wind speed (m/h)K = coefficient of effectiveness which is dependent upon the

direction of the wind relative to the opening and thearea ratio of the two openings (Bansal 139)

The rate of air flow through the building, therefore, increases withlarger openings, higher wind speeds, and perpendicular orientationof openings to incoming wind. When seeking ventilation with wind,cross ventilation is also desirable. Single sided ventilation will onlyprovide air movement to a very shallow depth of the building. Analternative is to provide an exhaust for the air via a ridge terminal orchimney. These conditions are diagrammed in figure 2-3. Addition-ally, the building should also be sited such that it will intercept thewind in the prevailing season. Figure 2-3: Opening configuration for

wind induced ventilation (Bansal: 145).

3.0 Development of the building design in relation to natural ventilation systemFrom the beginning of the project, Piano desired the incorporation of a natural ventilation system. Addition-ally, the engineering firm, Ove Arup, stressed the benefits of low technology because of the expense ofimporting and maintaining mechanical equipment (Banfi 26). This system, however, did not generate theinitial form of the building but rather responded to it. As the design progressed, the natural ventilation systemwas further incorporated into the design and exerted formal changes upon the building.

3.1 Competition entryIn the original scheme, the cases were found onboth sides of the circulation spine as shown in theplan of figure 3-1. These cases were inspired bythe traditional huts of the Kanak people (see figures3-2, 3-3 and 3-4). In fact, in this initial scheme, thecases were arranged into distinct groups whichwere thought of as “villages”. These cases alsosought to establish a relationship with the surround-ing vegetation, especially the tall evergreen treesas shown in figures 3-5 and 3-6 (Buchanan 192,Zabalbeascoa 4). Piano even went as far as pro-posing that local materials be used for the claddingand replaced periodically replaced by the commu-nity (Buchanan 192).

Figure 3-1: Plan of competition entry with cases on both sides ofthe circulation spine (Buchanan:199).

Figure 3-2: Drawing of the Kanak huts(Buchanan: 196.)

Figure 3-3: Sectional drawing of thecompetition entry showing the originalform of the cases (Buchanan: 199)

Figure 3-4: Photgraph showing relationshipbetween the forms of the final cases and theKanak huts (Findley: 98)

Figure 3-5 (left): The evergreen trees of Noumea (Buchanan: 195)Figure 3-6 (top): Sectional drawing of the competition entry showing the relationship to the tall trees(Buchanan: 198)

Two methods of natural ventilation were proposed in this competition entry. The first relied upon the place-ment of the cases on both sides of the promenade, some open in the direction of the wind and some withtheir back to the wind, to achieve an even ventilation of the building (Buchanan 192). This scheme utilizesthe principle of ventilation due to wind forces with the cases providing the openings on the windward andleeward sides of the building . Therefore, this ventilation scheme is not related to the tall, curved forms ofthe cases, but could be performed by a low, orthogonal building. The second proposed ventilation methodutilized the form of the cases as wind scoops. This, however, did not give the ventilation scheme enoughflexibility to respond to the different wind directions and strengths (Banfi 26), and wind tunnel tests demon-strated that wind was not being brought into the building (Buchanan 197).

3.2 Abstraction of the cases and incorporation of natural ventilationIn developing the building from this initial competition entry, changes were made to create a more feasiblenatural ventilation system. Some of these changes stemmed from formal design considerations. In otherinstances, introduction of ventilation elements exerted formal changes upon the building.

By January of 1992, there was an effort to make the cases more abstract, rather than quoting so directly fromthe Kanak vernacular architecture. Iroko wood, with its great durability, was chosen for the cases, rather thanopting for the frequent replacement of local materials. The vertical elements no longer meet at a point andwere no longer the same length. This new form allowed the ventilation of the cases to be converted from theunsuccessful wind scoops of the initial design into thermal chimneys. The north side of the cases (the sidetoward the promenade) were opened, and air could move through the cases and up the chimney. For afurther discussion of the thermal chimney, refer to section 3.3.

Additionally, during this design iteration, all of the caseswere moved to the south side facing towards the windward,ocean side (see figure 3-7) This allowed a structural systemto be designed to withstand the hurricane winds coming offof the ocean. It also had implications for the ventilationsystem, as all the cases were now facing in the windwarddirection.

By January 1993, the walls of the cases had developed intotwo concentric rings (Buchanan 199). The interior ring wascomposed of vertical columns of laminate iroko wood andformed the interior wall of the cases. The exterior ring wascomposed of curved laminated wooden members. Steelbracing and connections were used to connect the two ringsand make them rigid. This double wall construction greatlyimproved the performance of the proposed thermal chimney(Buchanan 199, Banfi 28).

Figure 3-7: Plan showing all the cases moved to thesouthern side facing the ocean (Buchanan: 205)

The performance of the ventilation system was considered further in the design of the cladding systems. Theexternal ring was clad in wooden slats from top to bottom which were placed more closely together at themiddle in order to reduce the wind at that point. This would allow wind to either pass freely through the top ofthe case, or to be forced downward to aid with the internal ventilation. Bringing wind in through a low openingprovides better circulation for the interior because air moves past the occupants rather than remaining atceiling height. These circulation patterns were also considered for the louver system that was introduced tothe internal ring. Louvers were placed at the base of the room to allow wind into the building. Anotherlouvered opening was placed at the ceiling to allow the heated air to escape up the thermal chimney. Theselouvered openings, which will be discussed further in section 3.3, are used to control ventilation through windforces and convection.

This proposed ventilation system was tested in the wind tunnel using 1/50 models with the blowers directedat the model’s outer face. Again sensors within the model registered no significant air movement enteringthe interior space through the openings in the shells. The CSTB (Scientific and Technical Building Center)engineers reasoned that an additional opening would draw air through the space and out. By cutting a holein the roof, they were able to achieve the desired ventilation. Rather than penetrating the roof, Piano addedsmall interior patios across the central walkway to induce the desired cross-ventilation (Hart 155,156).

Figure 3-8: Final plan of the cultural center (Findley: 98)

The need for the patios is again autilization of the principle of windventilation as discussed in section 2.2.These patios (or the proposed open-ings in the roof) allow for cross-ventila-tion to occur. Hence, the ventilationsystem exerted a formal change uponthe building by introducing theseinterior courtyards (shown in yellow infigure 3-8).

3.3 Analysis of natural ventilation system in final designThe final natural ventilation system incorporates both the principles of the stacked ventilation and of ventilationdue to wind forces. Depending upon the wind forces, different types of ventilation are utilized through theopening and closing of louvered apertures. These computerized louvers are located in the interior ring ofthe cases. At the ceiling are 2 meter high operable louver windows, and at the base there are 0.5 operablelouvers. Additionally, across the circulation spine from each case there is another computerized louveredopening at the courtyards. These computerized louvers respond to different wind speeds to control theventilation of the building to a maximum speed of 1.5 m/s (Banfi 26) .

The first mode, shown in figure 3-9, is usedwhen there are light winds or still air. With thethermal chimney closed, ventilation of thebuilding is solely dependent upon wind forces.The height of the building is not utilized toventilate the room.

Figure 3-9: Mode 1 for ventilation (Banfi: 27)

Figure 3-10: Mode 2 for ventilation (Banfi: 27)

The second mode, shown in figure 3-10, is usedwhen there are moderate winds and lightbreezes). Examining the basic equation forventilation due to wind forces:

Q = KAV(refer to section 2.2) it can be seen that with thehigher wind velocity, the smaller opening willmaintain the same rate of airflow as in mode 1.Again, ventilation relies solely upon wind forces,and the height of the building is not utilized.

The third and fourth modes, as shown in figure3-11, are used when there are strong winds. Thismethod relies upon both convection and thenegative pressure created by the wind. Utilizingthe principles of stacked ventilation, the airentering the room is heated and exits throughthe upper opening and up the thermal chimney.This is aided by the fact that the wind is creatingnegative pressure across the top of the thermalchimney, causing the air to be sucked up thechimney.

During cyclone conditions, as shown in figure 3-12, all of the louvered opening are closed.

There are times that the winds come across thelagoon, rather than across the bay, as shown infigure 3-13. During these times, both theprinciples of wind ventilation and stackedventilation are used. The louvers on both thesides are opened, allowing the wind to movefrom the positive pressure on the northern sideto the negative pressure on the positive side.Additionally, as air heats up within the room, itrises through convection and can exit out of thethermal chimney.

Figure 3-11: Modes 3 and 4 for ventilation (Banfi 27).

Figure 3-12: Ventilation system during hurrican (Banfi 27).

Figure 3-13: Ventilation system during reverse winds (Banfi 27).

Ove Arup calculated the percentage oftime each mode would be used andfound that mode 3, which utilizes bothprinciples of ventilation, would dominate(see figure 3-14). Given that mode 1would hardly ever be in operation, theselouvers were made to be manuallyoperated (Banfi 27). Ove Arup per-formed a comfort analysis to see if thisventilation system would be successful.Air temperature, radiant temperature, andhumidity were calculated for each hourevery day of the year at several points inthe occupied space. Using data fromthe wind tunnel tests, it was possible toapproximate the internal air velocityunder each mode of operation.By combining this data, a comfort index for each hour of the critical months could be calculated. A compari-son with the Gagge acceptable comfort criteria showed that during the hottest month, February, the criteriawere exceeded for only 5.8% of the occupied hours (Banfi 27). This was deemed acceptable by the client,and upon completion in autumn 1998, comfortable conditions had been achieved in the naturally ventilatedspaces (Banfi 29).

Figure 3-14: Comfort analysis data (Banfi: 27)

4.0 ConclusionThe success of the ventilation system of Renzo Piano’s cultural center is dependent upon the scale at whichit is examined. Although it is quite successful when one examines the design of an individual case, severalcritiques can be made at the scale of the overall building system.

The first such criticism is that the repetitiveness of the system. There is a lack of consideration for the indi-vidual environment of the cases. For example, the computer lab housed in one of the cases would generatemuch more heat than an exhibition space. Yet, with all the cases having the same openings and the comput-erized louver system controlled by the exterior wind pressure, the ventilation cannot be adjusted to take thisinto account. Secondly, this ventilation system does not function throughout the building. There are airconditioned spaces within this building, although the published materials barely make mention of this fact. Insome ways this reduces the cases, with their computerized louver system, to an iconography of a naturalsystem that does not exist throughout the building.

On the scale of the individual cases, however, the natural ventilation system functions successfully. Althoughthe thermodynamic principles used did not initially generate the form of the cases, the adaptation of their formto these systems is critical. By allowing the cases to perform this functional role, and allowing the ventilationsystem to have an impact on the form of the building, the cases are no longer iconographic symbols of theKanak culture and environment.