Cholera Treatment Center in Haiti

July 9, 2012 / no comments

Recently we were asked by MASS Design Group to do a daylighting study for their new Cholera Treatment Center (CTC) in Port-au-Prince Haiti. This is the second time that we have had the opportunity to work with them. Since the earthquake on January 12, 2010, Haiti has suffered enormous economic, structural and environmental distress. Cholera, which previously did not exist in Haiti, broke out shortly after the earthquake and according to the World Health Organization, as of November 30, 2011, there were 515,699 reported cases of cholera and 6,942 deaths.i The new CTC will be a year-round center for the care and treatment of patients suffering from the illness. This project presents an unusual opportunity to engage lighting on such fundamental levels, and to think about basic human needs as they relate to light. It’s even more unusual to be facing these issues in a country like Haiti. As designers in the United States, our typical projects include commercial, residential and institutional projects in major developed cities. While many basic principles of good lighting remain universal, working on a project like this exposes us to significant cultural, financial and climatic differences.

What we learned

MASS Design Group is a non-profit architecture and design firm dedicated to helping improve the health and overall well-being of communities through design. Their most well-known and highly-praised projects include the Butaro Hospital and the Girubuntu School, both in Rwanda.

During our design conversations, members of the MASS team shared with us examples of the kinds of design and technology challenges that they have encountered in their work. One team member, Elizabeth Timme, shared a few experiences from Butaro Hospital. The hospital is tailored for the treatment of tuberculosis, so enormous fans are suspended from the ceiling and UV lights shine from the walls. The fans keep air moving (a critical element in tuberculosis treatment), and the UV lights kill airborne bacteria. Patients however, worried that the fans were “stealing their air” and that the lights were “burning their skin”. To some extent both comments are true, but negotiating the relationship between the function of the space and the occupant use and comfort highlights a fundamental mission of design.

For decades Western hospitals were built to enhance efficiency and hygiene, and only recently has there been more evidence that daylight, views and good design are equally critical factors to providing the best health care. In fact, Timme is now starting her own firm, Más, in order to bring a similar design approach as used at MASS to the issues facing the American health care system.

MASS has encountered other issues that have to do with climate and resources. In one case, they told us that in Rwanda, surface brightness was a serious design consideration. The sun at that latitude can cause surfaces to be so bright that they create visual discomfort. An article by Martin Schwartz about Louis Kahn’s proposal for a U.S. Consulate in Luanda illustrates this issue. In it Schwartz quotes Kahn:

“I …noticed that when people worked in the sun-and many of them did-the native population …usually faced the wall and not the open country or the open street. Indoors, they would turn their chair toward the wall and do whatever they were doing by getting the light indirectly from the wall.” ii

Not only do these types of realizations help us to understand better how to work in under-served countries, but they help to inform our approach to design for our typical projects. Working on a cholera treatment center can help us to recognize and consider factors in our day-to-day work that we may not have previously considered to be important or necessary.

What we did

The center is 7,700 square feet and consists of a general ward, an intensive care ward, as well as an administration office. The roof consists of 15 roof modules, each 23′ x 23′ oriented in different directions. Four central modules serve as the top of a cistern designed to collect water, while the remaining 11 are pitched, both to allow for light and ventilation as well as to direct water towards the cisterns. The building is primarily open on all sides, although there are fixed screens on the North and East sides and sliding screens hung on the West side.

Since construction had already begun when we joined the team, the project manager, David Saladik, emphasized that we needed to quickly and efficiently produce a series of clear and useful daylighting studies. There was no possibility for bells or whistles. We had to make simple and effective recommendations that could be executed within the time and resources available. It was important to set our criteria immediately, and use tools that would help us achieve our goals. For all of the simulations we used the daylighting program DIVA-for-Rhino.

Our analysis focused on three key issues: sufficient light levels, glare, and heat gain. To test for adequate light levels, we ran a daylight autonomy test using a horizontal calculation grid at the height of the beds, and vertical grids at the headboards. We used a threshold of 500 lux because light levels need to be relatively high, since nurses and doctors will be conducting procedures and treating patients at their bedside. We were pleased to see that the daylight autonomy results for the existing design showed that we can expect 500 lux over the majority of the ward space for 50% of the year.

1_Blog CTC DA 500 plan.jpg

Daylight Autonomy calculation (500 lux threshold)

Having established a general sense for annual daylight levels, we wanted to look more closely at the lighting conditions during specific points in time when we could estimate there might be problems. The East side is most vulnerable to overlighting and heat gain because there are no structures directly to the East of the center. We wanted to test a variety of material options for a proposed set of panels, which would serve as shading. We chose three materials to test: opaque panel with 50% reflectance, a translucent panel with 40% transmission, and 40% open screen.

2_Blog CTC Screen Analysis.jpg

Evalglare Glare Analysis

Summer Solstice 9am Glare Analysis 40% Open Screen (top right)

Winter Solstice 9am Glare Analysis of (3) different panel options (bottom)

Our results show that only the opaque panels will have a significant effect on minimizing the overlighting on the East-side ward. The translucent and 40% open screen still allow a significant amount of light and, by extension, heat into the building. In addition, when we ran the glare tests for the three material options, the 40% open screen performs the worst in terms of glare for the patients in East-facing beds. This only gets worse during the rest of the year when the sun is lower in the sky; on the Winter Solstice at 9am, the Evalglare Daylight Glare. iii Probability result showed there would be intolerable glare . Since the preferred solution is to use the screens, we recommended modifying the open percentage of the screen to be less than 40%.

3_CTC Roof Analysis.jpg

Glare experienced from hospital bed on Summer Solstice at Noon (bottom left)

Roof module and assembly (bottom right)

Cholera outbreaks in Haiti are the most frequent during the rainy season, which happens in the summer. Knowing that, we were particularly interested in the summer conditions since those are the times the wards will likely be most full. In particular, we looked for problems with glare and overlighting. The orientation of the roof modules on the East side allowed, rather than prevented, direct early morning and noontime sun to stream into the building. We confirmed this by running several glare simulations, and made the recommendation that two of the roof modules be rotated 90 degrees in either direction to eliminate the problem.

4_Blog CTC Trough Analysis.jpg

Illuminance Calculations Summer Solstice 12pm

Existing Glazing Design (left). Glazing replaced with 80% Opaque Panels (right)

The last issue we wanted to tackle was the glazed portions of the roof that would direct water into the cisterns. Those sections would be glazed in order to let light into the center of the building, and be directly above a small, planted garden. We ran several tests on the Summer Solstice, which showed that there would be dramatic overlighting, glare and potentially great heat gain. While we understood the design intent to create a small lush area in the middle of the ward, we suggested some simple louvers to mitigate the quantities of incoming daylight.

While absent of high-technology or complicated details, our solutions answered the key daylighting questions and will contribute to providing the patients and staff with the most functional and comfortable space possible. Even with the demanding constraints of a project like this, in the end, we’ve shown that with targeted thinking and the right tools, we can come to useful conclusions and make effective recommendations, which is definitely a lesson we can bring back to our everyday work.


i World Health Organization, “Health Cluster Bulletin: Cholera and Post-Earthquake Response in Haiti”, 21 December 2011.

ii Louis I. Kahn Writings, Lectures, Interviews, page 123, quoted by Martin Schwartz in “Louis I. Kahn: Finding Daylight in Luanda”, February 7, 2011.

iii Evalglare calculates glare according to several factors such as brightness and size of glare source, and gives a result called the Daylight Glare Probability, which was calibrated by user-assessments. The categories of glare perception from least severe to most are: Imperceptible, Perceptible, Disturbing and Intolerable. For more information see: Evalglare.

Image Credits: Kera Lagios/Lam Partners

Choosing Glass for Daylighting

July 5, 2012 / no comments

Glazing for daylighting needs to be chosen both for its visual character (clear, diffusing, etc.) and its technical light-transmission values. Commonly, the visual decision will come first. Do we want clear glass, complete diffusion, or something in between? Diffusing glass has its applications, but needs to be used with care. While it’s tempting to “solve” the problem of diffusing sunlight and controlling direct sun glare by using diffuse glazing, that glazing itself can also become a source of glare. Diffusing glazing with high light transmission, such as Kalwall, can be dazzlingly bright if it’s exposed to direct sun. If that’s in our field of view it can be a worse source of glare than the sun through clear glass because, unlike clear glass, it’s bright from all viewing angles. And of course diffuse glazing destroys outside views, which can be as valuable for well-being and productivity as the daylight itself. For these reasons, fully diffusing glazing is best used where it is not in direct view. Opal frits and silkscreened interlayers can provide a combination of diffusion with some clear views, however they may still become glare sources, similar to fully diffusing glazing. So let’s assume we’ve chosen clear glass (whether tinted or not) for its visual character – now what light-transmission values do we want?

All practical sources of illumination contain a mixture of visible light energy and non-visible energy. What happens to the energy in daylight after it enters an interior space? Assuming that our space is not open to the outdoors, typically only a small amount of the daylight will be bounced back through the apertures to the outside. A very tiny amount of the daylight energy will break up molecules in our fabrics, artwork and other fragile materials. The remaining daylight energy will all be absorbed in interior surfaces and converted into heat. Much of the time, especially in large buildings, that heat is undesirable and will need to be pumped out of the space by a mechanical system, consuming energy. Our choice of glazing can have a major impact on how much heat gain results from daylight.

No glass is totally transparent – only a fraction of the light that hits it passes through. The good news is that readily available glass types admit daylight selectively. In other words, the fraction of visible light that is transmitted is often different from the fraction of total light energy that is transmitted. So if we choose our glazing well, we can admit less total energy (heat gain) for a given amount of visible light than we would if there were no glass, just an opening.

A useful way to evaluate glazing is to look at the ratio of visible light transmitted to heat gain transmitted. A good measure of the heat gain is the SHGC (solar heat gain coefficient). This information is readily available from major glass manufacturers, and some of them even calculate the ratio for us. Viracon, for example, calls this the LSG (light to solar gain ratio), and provides it in their glass data charts. For example, their data chart for 1-inch low-E insulating glass with argon fill shows that for clear glass (VE 1-2M) we can get a visible transmittance of 70% and an SHGC of 0.37, giving a very favorable ratio (LSG) of 1.90. So using this glass will cut the heat gain nearly in half for a given amount of admitted visible daylight. If we use glass tinted blue-green, the ratio can even be a bit higher, 2.01, however that improvement may not be worth the resulting coloration of the daylight.

An interesting way to think about this is that we can say we are using selective glazing to actually improve the energy quality of the daylight itself. The technical term for this quality is “efficacy”, which can be applied to artificial light sources as well as to daylight. Conventionally, the visible light energy is measured in lumens and the total (heat gain) energy is measured in watts. So efficacy is expressed as lumens per watt. The efficacy of outdoor “raw” daylight from sun and sky combined varies with sky condition and solar altitude, but generally runs around 100 to 120 lumens per watt. So for one watt of heat gain we get 100 to 120 lumens of visible light. Interestingly, this is very comparable to efficient electric light sources. With a high-efficiency T8 fluorescent lamp, for example, it will take just about 1 watt (including the energy to operate the ballast) to produce 100 lumens. Large HID lamps (metal-halide and high-pressure sodium) can produce more than 100 lumens for each input watt. So, other things being equal, raw daylight and electric light would result in the same heat gain to our space. But once we have passed the daylight through selective glazing, we can multiply its efficacy by the LSG ratio, so for the clear glass example it’s now around 190 to 230 lumens per watt. This cuts the daylight heat gain more or less in half. The daylight would now also create half the heat gain of an equal amount of electric light. The key word there is “equal”, since in practice daylight levels in a space will often be much higher than electric lighting levels, with resulting higher heat gain (see Daylighting Reduces Heat Gain – Pantheon Redesign?).

If a design mandates a certain amount of glazing, we can adjust the daylight levels by choosing different visible light transmissions. For example, the designer may want to choose a lower visible-transmission glass if the amount of glazing in the design creates daylight levels which are higher than necessary. Choosing glass with a high LSG ratio is still desirable in that case, but as the Viracon data shows, the ratio will be lower than with high-transmission glass. So that design will be paying a penalty in heat gain from daylight (and probably from the U-value of the extra glass also), compared to a design using a smaller area of high-transmission glass to accomplish the same daylight levels.

In some cases, a design could even benefit from different glasses at different apertures, for example, a high-transmission glass at a small clerestory and a lower-transmission glass at a large lower window. This approach should be used with caution – when all of the glass is the same, we don’t perceive the glass transmission very clearly. But when two different glasses are visible from the same viewpoint, the lower-transmission glass can look gloomy by comparison.