Responsive Daylight Control: The “Mount Everest” of Automated Window Shading

Responsive daylight control, sometimes called dynamic daylight control, is the capability to automatically adjust window shading to maintain a desired level of admitted daylight. This is arguably the most valuable of the five levels of shading automation, but also the most challenging to implement cost-effectively. In fact, responsive daylight control is virtually non-existent in U.S. residential or commercial buildings—and the reason is that current technology falls short in both effectiveness and affordability.

Responsive Daylight Control versus Automatic Glare Blocking

Responsive daylight control is simple in concept: if there’s not enough daylight, open the shading until there is; if there’s too much daylight, close the shading until there isn’t.

However, virtually none of today’s automated shading products that claim to respond to daylight actually provide responsive daylight control as described above. Instead, they attempt to provide what we call automatic glare-blocking, which is supposed to fully close the shading when there’s a risk of glare, and then open it to the previous setting when there isn’t.

In the parlance of control theory, responsive daylight control provides proportional (continuous) control, while automatic glare blocking provides bang-bang (discontinuous) control. Both provide the same type of benefits, but responsive daylight control provides them to a greater degree. The following charts show how responsive daylight control and automatic glare-blocking respond to the variation in irradiance over a cloudless sunny day:

Automatic glare blocking keeps the shade fully open until the irradiance exceeds a threshold, and then fully closes it; responsive daylight control closes and opens it just enough to keep the admitted daylight close to a desired daylight set-point. As a result, the average daily level of admitted glare-free daylight is much greater with responsive daylight control than with automatic glare-blocking.

On the other hand, responsive daylight control is also harder to implement. For example, it requires more frequent and more subtle shading adjustments, which is practical only with certain types of shading device. The differences between responsive daylight control and automatic glare blocking are summarized in the following chart:

While this post is mainly about responsive daylight control, automatic glare blocking presents many of the same qualitative benefits and challenges, although to a lesser degree.

The Benefits of Responsive Daylight Control

People generally prefer daylight to artificial light, but they also can’t tolerate the glare that comes from too much daylight. That’s why windows are typically equipped with manually operated window coverings. Responsive daylight control wouldn’t be very useful if people regularly adjusted their shading throughout the day.

However, manually operated window coverings actually aren’t operated very often. In fact, their average adjustment frequency is less than one per day, and they’re typically left in a mostly closed position—and that applies to both residential and commercial buildings.

For example, per the research cited in the above-linked posts, the median setting for horizontal blinds on sunny windows in office buildings results in 70% window occlusion (i.e. with the slats lowered to cover 70% of the window area) and a 40-degree slat tilt (to mostly block a sky view), resulting in a relative daylight transmittance of only about 0.4:

The fact that window coverings are mostly closed makes sense because people have better things to do with their time than to constantly adjust them—and the potential repercussions of an under-shaded window (excessive heat and glare) are worse than those of an over-shaded window (reduced natural illumination and a blocked view).

But this also means that windows are almost never optimally shaded:

• Most of the time, a window with a mostly closed window covering will admit far less useful glare-free daylight, and provide a more obscured outward view, than it otherwise could.
• However, during periods of peak irradiance, the same window will admit more solar heat than if it were fully closed, unnecessarily increasing the cooling loads on the HVAC system.

That’s the problem solved by responsive daylight control. By automatically providing the right amount of shading under changing conditions, responsive daylight control can provide three significant benefits:

• Increased glare-free natural illumination. Not only do people generally prefer daylight to artificial light, there is also evidence that increased levels of glare-free daylight have significant positive effects on mood, health, and productivity. In fact, this is almost certainly the most valuable benefit of responsive daylight control, but also the most difficult to quantify.
• Energy savings via reduced need for artificial lighting. The increased natural illumination can be “harvested” in the form of energy savings by turning off the lights when they aren’t needed. This could theoretically be done manually, but for maximum savings should be done automatically using an inexpensive daylight-harvesting lighting control.
• Energy savings via reduced cooling loads on the HVAC system. This is a natural byproduct of responsive daylight control and needs no intervention by the lighting or HVAC systems, but is obviously significant only in climates where HVAC cooling is sometimes necessary.

Our market research shows that these benefits resonate differently in the non-residential and residential market segments.

Benefits of Responsive Daylight Control in Non-Residential Buildings

There is plenty of research showing that responsive daylight control can provide compelling benefits in non-residential buildings.

Non-Energy Benefits

Most buildings (other than warehouses) exist for the sake of the human activities they house. Nevertheless, investments in building technologies are usually made primarily on the basis of building operating costs, and only secondarily on the potential for positive impacts on the building occupants.

That’s unfortunate for responsive daylight control, because it’s been proven that increased glare-free daylight can increase productivity in office buildings, increase gross sales in retail stores, and improve test scores in schools—but that too much daylight (in the form of glare) has a negative impact on the same metrics.

Summaries of some of research on the positive ergonomic impacts of daylight can be found at the websites of the Velux Group and the Rensselaer Polytechnic Institute’s Lighting Research Center.

But probably the most-cited and thorough studies on the effects of daylight on human performance in non-residential buildings were those performed by the Heschong-Mahone Group in the late 1990s and early 2000s. These studies are particularly useful because they account for the differences between top-lighting with skylights (which generally yields glare-free daylight) and side-lighting with view windows (which can often cause daylight glare). Here are some tidbits from those studies:

• A study of 100 workers in a call-center found that they were able to process calls 6% to 12% faster when they had the best possible outward view, versus no view (Heschong-Mahone Group, “Windows and Offices…,” Executive Summary, page vii).
• A study of 200 office workers found that they scored 10% to 25% better on tests of mental function and memory recall when they had the best possible outward view, versus no view. However, in some of the tests, performance decreased by 15% to 21% when the outward view was accompanied by occasional glare (Ibid.).
• A study of test score results for over 21,000 elementary students in three school districts showed that rates of learning are strongly correlated with daylight and outward view. In one school district, students with the most daylight in their classrooms progressed 20% faster on math tests and 26% on reading tests than those with the least daylight, and students with the largest window areas progressed 15% faster in math and 23% faster in reading than those with the smallest window areas. In two other school districts, students in classrooms with the most daylight were found to have 7% to 18% higher end-of-year test scores than those with the least daylight. (Heschong-Mahone Group, “Daylighting in Schools…,” Executive Summary, pp. 2 – 3).
• A study of 108 retail stores found that skylights were positively and significantly correlated with higher sales, and suggested that an average non-skylit store in the chain would likely have 40% higher sales with the addition of skylights. In fact, after the number of hours open per week, the presence of skylights was found to be the best predictor of the sales per store of all the variables considered. (Heschong-Mahone Group, “Skylighting and Retail Sales.” Executive Summary, p. 2)
• Because of the surprisingly large association between sales and skylights found in the previous study, a more detailed follow-on study was performed in different stores. This found a much smaller but still significant (0% to 6%) increase in sales due to daylight, and that daylight was found to have as much explanatory power in predicting sales as other more traditional measures of retail potential, such as parking area, number of local competitors, and neighborhood demographics. (Heschong-Mahone Group, “Daylight and Retail Sales.” Executive Summary, p. vi).

These findings show that daylight and outward views have a significant positive impact on human activity if the daylight is glare-free, but also a significant negative impact if there is daylight glare.

To put these findings in perspective, a change of even a few percent in a metric related to human activity (such as test scores or annual retail sales) typically represents far more economic value than the energy costs of operating a building. For example, consider a typical office building:

• Electricity expenditures will typically be less than about $2 per square foot per year.
• In contrast, assuming an average annual salary of $60K and an area allocation of 160 square feet per employee, the annual salary expenditure would be $375 per square foot—fully two orders of magnitude greater than the energy costs.

In this particular example, a mere 0.5% improvement in human productivity would be as valuable as a 100% savings in electricity costs.

Energy Benefits

Despite the fact that the positive impact on occupant activities is almost certainly more valuable, the potential for energy savings is what has driven most of the interest in responsive daylight control for non-residential buildings—and for office buildings, in particular.

As previously mentioned, responsive daylight control saves energy in two ways:

• It reduces cooling loads on the HVAC system by fully closing the shading when the incident irradiance is strongest. Of course, this is relevant only to climates with a significant number of cooling degree-days, but that represents a substantial fraction of the built environment.
• It reduces the need for artificial lighting by opening the shading when there is no risk of glare. This allows the lamps to be dimmed or turned off to save lighting energy, which is referred to as daylight harvesting.

Fully Exploiting the Energy Savings Entails Daylight Harvesting

Daylight harvesting is a component of a broader architectural strategy known as daylighting, which is aimed at maximizing the ratio of natural to artificial illumination. Daylight harvesting involves dimming or turning of the lights when there’s enough natural light to meet the occupants’ illumination needs. It can be done manually, but is ideally done automatically using a daylight-harvesting lighting control which automatically adjusts the artificial lighting level to maintain a desired level of total (artificial plus natural) illumination. Thus, any increase in daylight is “harvested” in the form of reduced power consumption in the lighting system.

Inexpensive daylight-harvesting lighting controls have been available for decades, but they’re used in fewer than 2% of U.S. office buildings (more on that statistic later in this post). One reason is that they have the reputation of often failing to provide the expected savings—and the culprit appears to be the chronic over-shading associated with manually operated shades.

According to a 2005 study—perhaps the most comprehensive study ever of the actual, real-world savings provided by daylight-harvesting lighting controls in side-lit spaces—the average relative savings in lighting energy were 3.59 Full-Load-Hours (FLH) in side-lit spaces without manually operated blinds, but only 1.84 FLH in spaces with manually operated blinds (“Sidelighting Photocontrols Field Study” 2005, Appendix F, Table of Full Load Hour Savings, Category “window controls”). Thus, the spaces with blinds were providing only about 50% of the energy savings of the spaces without blinds. Further, these statistics actually imply a much greater savings loss due to over-shading, because only the windows that received little sunlight were likely to be without blinds.

Note that the approximately 50% lower savings in areas with blinds is consistent with the median 40% relative blind transmittance previously shown in Figure 4.

This is pretty strong circumstantial evidence that over-shaded windows are responsible for the relatively low savings from daylighting in areas which are side-lit with ordinary view windows—and that responsive daylight control has the potential to dramatically increase the savings.

Responsive Daylight Control Recovers the Savings Lost Through Over-Shading

It should be obvious from the discussion so far that responsive daylight control can save at least some energy. Exactly how much energy can be saved depends on numerous site-specific factors, but studies have shown that the total savings can be a substantial fraction of the total (HVAC plus lighting) energy consumed in windowed areas. In fact, responsive daylight control has been considered so promising for energy savings that the Department of Energy has sponsored its development, off and on, for decades.

The pioneering research in this field was done at Lawrence Berkeley National Laboratory (LBNL) in the late 1990’s. One LBNL study—the first ever involving full-scale testing of a prototype automated venetian blind in actual (but unoccupied) offices in Oakland, CA—confirmed that responsive daylight control could substantially increase the savings from daylight harvesting when compared to manually operated blinds (Lee et al., 1998). The following chart shows the cooling and lighting savings observed for three different slat-tilt angles of the base-case blind (i.e. the blind without responsive daylight control):

As shown in the chart, the lighting savings increased and the cooling savings decreased with the amount of base-case shading—but the total savings over a manually operated blind were significant regardless of the base-case shading. From this Figure 6 data, the median base-case slat tilt angle of 40 degrees previously shown in Figure 4 would result in lighting savings of about 30% and cooling savings of about 12% relative to a manually operated blind with the same daylight-harvesting system.

Those are impressive numbers, but they represent relative savings, not absolute savings in kWh. The lighting system in LBNL’s late-1990s study used fluorescent lamps with a Lighting Power Density (LPD) of 1.35 Watts per ft²; today, thanks to more efficient LED lighting, the LPD in office buildings averages only about 0.8 W/f² and is sometimes as low 0.35 W/f². So, the lighting savings in kWh would be much less today than in the late 1990’s.

In some areas (notably California, the site of LBNL’s testing), the reduction in LPD has been accompanied by dramatic increases in retail electricity prices, so that the value of the energy savings is roughly the same today as it was back then. However, the nationwide average retail price for electricity in the commercial sector has risen by only 40%.

The upshot is that a system like the one prototyped by LBNL will be less cost-effective for energy savings today than it was in the 1990’s, except in areas with the highest electricity prices (e.g. California, Alaska, Hawaii, and the northeast U.S.). The obvious question is whether it would still be cost-effective enough to appeal to mainstream purchasers—and we’ll get to that later in this post.

Benefits of Responsive Daylight Control in Residential Buildings

While there is plenty of research on the benefits of responsive daylight control in non-residential buildings, we haven’t found any research that specifically addresses residential buildings.

However, there’s no question that responsive daylight control can’t save as much energy in residential buildings as in commercial buildings: residential buildings have a lower lighting energy intensity (kWh per square foot) and a smaller ratio of window area to wall area—and those are the two first-order variables that determine the achievable savings.

On the other hand, our own market survey shows that the non-energy benefits of responsive daylight control resonate strongly with potential residential purchasers, and much more strongly than in the commercial market:

• The potential health and wellness benefits of increased glare-free daylight are extremely appealing to many potential residential purchasers. Many people are already aware of the link between insufficient exposure to daylight and health issues such as Seasonal Affective Disorder and disrupted sleep patterns, and are willing to accept the premise that responsive daylight control can be helpful in mitigating these issues.
• The promise of increased glare-free natural light per se has significant appeal for most potential residential purchasers, independent of any potential health or wellness benefits.

Our survey shows that the potential for energy savings is also appealing to many potential residential customers of automated shading products in general, who can be grouped into two categories.

• Potential purchasers who are interested primarily in energy savings. These purchasers typically demand a payback period shorter than is possible with responsive daylight control.
• Purchasers who are interested primarily in the non-energy benefits of automated shading. The energy-saving potential of responsive daylight control resonates as a significant secondary benefit with these purchasers, even without a quantified (or short) payback period.

Overall, our market survey suggests that the benefits of responsive daylight control are compelling enough to elicit significant mainstream demand, but—as with the commercial market—only if the benefits can be provided at a low enough price. More on that later.

Despite Compelling Benefits, Responsive Daylight Control Remains Extremely Rare

Despite a compelling value proposition, actual use of responsive daylight control appears to be virtually non-existent in either residential or commercial buildings.

Use of Responsive Daylight Control in Residential Buildings

There are no independent statistics on the use of any form of automatic daylight control in residential buildings.

However, a survey of online descriptions of residential automated shading products suggests that fewer than 20% of products claim an out-of-the-box capability to automatically respond to daylight levels. Further, those products appear to promise only some form of automatic glare blocking, and not responsive daylight control as described above.

Virtually all residential automated shading products can be integrated with a home-automation platform, and some home-automation platforms do offer the possibility of automatic glare blocking (but not responsive daylight control) by adjusting the shading based on the calculated position of the sun. However, an online search suggests that only the most technically savvy users attempt to take advantage of this feature, and often with poor results.

And, again, that’s only automatic glare blocking, and not responsive daylight control.

Use of Responsive Daylight Control in Office Buildings

As previously described, responsive daylight control is even more advantageous in office buildings than in residential buildings, so it would naturally be expected to be adopted more quickly and widely in office buildings.

Unfortunately, as with residential buildings, there are no credible independent statistics that directly address the use of responsive daylight control in office buildings.

However, there are credible statistics on the use of daylight harvesting. Daylight harvesting is less expensive and more mature than responsive daylight control or automated glare blocking, but is necessary to fully exploit the energy-saving potential of either capability. Therefore, any responsive daylight control or automatic glare blocking installation in an office building would almost certainly also include daylight harvesting.

Per the 2018 Commercial Buildings Energy Consumption Survey (the most recent CBECS), only about 1.6% of office buildings were equipped for daylight harvesting (statistic derived from data in “Table B11.  Selected principal building activity: part 1, number of buildings, 2018“). Therefore, the penetration of daylight control of any type couldn’t have been more than 1.6% of office buildings, and was probably a lot less. Our own estimate is that fewer than one out of ten daylight harvesting installations include any form of daylight control—and most of those are undoubtedly of the automatic glare blocking type, rather than true responsive daylight control. This would put the use of any form of daylight control at less than a tenth of a percent.

Could use of daylight harvesting (and daylight control) have grown significantly since 2018? It’s possible, but the 1.6% indicated in the 2018 CBECS was actually lower than the 2.8% indicated in the preceding (2012) CBECS, indicating that daylight harvesting had reached a plateau by the time the 2018 survey was conducted.

The reason for the paltry penetration of daylight harvesting is open to speculation. However, we believe a major factor is the impact of over-shaded windows on the achievable savings (and purchasers’ confidence in actually realizing those savings). By eliminating the over-shading, responsive daylight control could unlock the full potential of daylight harvesting, significantly increasing its market penetration.

What’s Holding Responsive Daylight Control Back?

So, with such compelling potential benefits, why is responsive daylight control (and even automatic glare blocking) virtually non-existent in today’s buildings? There are three reasons:

• Currently technology can’t provide a short enough payback period to be considered cost-effective in non-residential buildings.
• Current technology can’t deliver the benefits of responsive daylight control without undesirable side-effects in the form of erratic and potentially annoying operation.
• Current technology is too complex and requires specialized site-specific set-up.

Excessive Payback Period in Non-Residential Buildings

The metric most often used to evaluate the cost-effectiveness of energy saving investments is the simple payback period, which is the upfront cost divided by the value of the annual savings.

When energy-saving retrofits for commercial buildings first became a hot topic in the aftermath of the 1973 energy crisis, purchasers were typically demanding a payback period of 3 years or less. It didn’t make economic sense to demand such a short payback, because a 3-year payback represents a Return-On-Investment (ROI) of 33% with minimal risk—far, far better than virtually any other type of investment. Nevertheless, that’s how the market initially worked.

However, as might be expected, non-residential purchasers eventually did become more tolerant of longer payback periods. Today, it’s generally accepted that a projected median payback period of 10 years or less is short enough to appeal to enough purchasers to ensure commercial viability.

So, what are the price and payback period of responsive daylight control in non-residential buildings? That depends on the type of shading device.

Motorized Blinds versus Smart Windows

As shown way back in Figure 3, only two types of shading device are well-suited for responsive daylight control: motorized blinds (as used by LBNL in their pioneering study) and Smart Windows. Unlike other shading devices, both of these provide a continuously variable transmittance over the entire window area, which is critical for responsive daylight control.

Smart Windows are considered the future of responsive daylight control because they have no moving parts and operate completely unobtrusively. That’s why the U.S. Department of Energy has funded Smart Window development for decades, and why the Inflation Reduction Act (IRA) of 2022 includes substantial tax incentives for their use (and not for the use of motorized window coverings). Perhaps more significantly, electronically actuated Smart Windows that could be used for responsive daylight control have been on the market for years.

On the other hand, Smart Windows are about five times more expensive than typical motorized blinds. And surprisingly, they actually can’t save as much lighting energy as motorized blinds, although they are more effective at reducing HVAC cooling loads.

In fact, Smart Windows are so much more expensive than motorized blinds that they aren’t direct competitors. Instead, virtually all of today’s Smart Windows are packaged in Insulating Glass Units (IGUs) intended as alternatives to conventional high-performance window units (i.e. spectrally-selective low-e IGUs), which are themselves a lot more expensive than motorized blinds. Like conventional high-performance windows, Smart Window IGUs make sense only for new construction and major building renovations (where new or replacement windows are going to be purchased anyway), whereas motorized blinds potentially also make sense as energy-saving retrofits in operating buildings. The advantage of retrofits is that they can proliferate a new technology far more quickly than new construction or major renovations, since only a small fraction of the building stock is newly constructed or completely renovated each year.

Let’s look at the payback periods for today’s motorized blinds and Smart Windows in the scenarios in which they make the most sense, i.e. in retrofit and new construction applications, respectively.

Payback Period with Motorized Blinds

In their 1998 study, the LBNL researchers estimated a 10-year payback period for their blind-based system in the Oakland, CA area at a flat electricity rate of $0.09 per kWh (Lee et al., page 17). Notably, LBNL calculated the payback period in a such a way that it was valid for retrofit installations in existing buildings as well as for new construction.

Many things have changed since then—some that would tend to lengthen the payback, and some that would tend to shorten it:

• The average Lighting Power Density (LPD) in office buildings today is only 0.8 W/ft² versus the 1.35 W/ft² LPD in LBNL’s test installation, tending to lengthen the payback.
• Average nationwide electricity rates have increased by about 40% (although they have virtually tripled in some parts of the country), tending to shorten the payback.
• A 1998 dollar is worth almost $2 today, tending to lengthen the payback by inflating the cost of the hardware.
• LBNL assumed a base-case blind tilt angle of 15 degrees, whereas a more appropriate assumption (per Figure 3) would have been 40 degrees. Per Figure 6, the greater base-case tilt angle would have increased the savings and tended to shorten the payback.

Overall, these factors imply a nationwide-average payback today of about 1.7 times the payback estimated in 1998, or about 17 years instead of LBNL’s estimated 10 years.

However, that 10-year payback estimated by LBNL was for a complete integrated shading-lighting system that also included the components necessary for daylight harvesting. LBNL didn’t provide a payback estimate for just the motorized blind portion of the system, but data in the report suggests that it would have been about 25 years.

When multiplied the by 1.7 factor cited above, this implies a 2024 nationwide average payback of 42 years, and a price of about $12 per ft² of glazing, for a responsive daylight control system like the one prototyped by LBNL in 1998.

So, a system like LBNL’s prototype automated venetian blind certainly wouldn’t be considered cost-effective across the bulk of of today’s market. However, it would be on the verge of being cost-effective in areas with the highest electricity prices (specifically California, Alaska, Hawaii, and the northeast U.S.). Paybacks would be even shorter in the significant number of buildings still equipped with fluorescent lamps, but that’s a moot point because those buildings would likely be retrofitted with quicker-payback LED lighting before responsive daylight control is even seriously considered.

Payback Period with Smart Windows

There are two Smart Window technologies that are reasonably mature and suitable for the type of responsive daylight control we’ve been discussing: ElectroChromic (EC) technology and Suspended Particle Device (SPD) technology.

A detailed description of these technologies is beyond the scope of this post, but essentially EC provides adjustable transmittance by changing tint through a reversible electrochemical reaction (like charging and discharging a battery), while SPD works by altering the alignment of rod-shaped nanoparticles using a variable electric field.

SPD technology has been developed purely through the private sector. It seems to offer some compelling advantages over EC (such as faster response time), but there isn’t much technical information on SPD in the public domain, and there doesn’t yet appear to have been any independent testing of SPD windows for architectural applications.

On the other hand, EC technology has been the beneficiary of substantial government investment, is the only Smart Window technology incentivized in the 2022 IRA, and has been independently tested though government-sponsored research. Therefore, the following observations are based on EC technology, but we’ve found no evidence that they won’t hold true for SPD windows, too.

In 2017, the U.S. General Services Administration (GSA) evaluated EC Smart Windows in two office buildings (located in Sacramento, CA and Portland, OR) as part of their Green Proving Ground (GPG) program. GSA found that EC windows did not eliminate the need for interior window coverings, were not cost-effective, and presented the same challenges in balancing glare control with energy savings faced by motorized window coverings. Here are some quotes from that assessment (emphasis added):

• “At the GSA national average utility rate of $0.11/kWh and a mature market cost of $61/ft2 (as estimated by the manufacturer) and with the continued need for blinds and their associated costs, payback at the Portland test bed was estimated at 29 years. The incremental difference between installing EC windows and spectrally selective low-e windows was estimated at $37/ft2 with a payback of 13 years. “ (Ibid., “Limited Cost-Effectiveness…,” page 3)
• “EC hardware itself is generally mature. It is able to control glare and thermal discomfort and reduce lighting energy use and HVAC cooling loads. However, these test-bed evaluations found that it was challenging to satisfy occupants’ aesthetic and glare requirements, while implementing control strategies that delivered HVAC and lighting savings. For this reason, and because of its limited cost-effectiveness, widespread GSA adoption of EC windows in occupied general office space is not recommended at present.“ (Ibid., page 5)

In case it isn’t clear from the quote above, the EC windows were actually estimated to have a longer payback period than installation of spectrally selective low-e windows, so it isn’t surprising the GSA didn’t consider them cost-effective.

Also note that unlike LBNL’s payback estimate for their blind-based system, GSA’s payback estimates cited above are valid only for new construction (in which windows have to be purchased anyway) or major renovations (in which the existing windows are going to be replaced anyway, even if responsive daylight control weren’t going to be used). EC window paybacks in a retrofit scenario would be far, far longer.

Thus, GSA’s estimated 29 year payback for EC windows can’t directly be compared to LBNL’s 10-year estimated payback in 1998 (or our estimated 42-year payback in 2024) for motorized blinds. For more of an apples-to-apples comparison, consider the cited shading device costs:

• GSA assumed a “mature market” cost $61 per ft² of glazing for EC windows (current actual cost is about $100 per ft²).
• LBNL assumed about $7.50 per ft² of glazing (in 1998 dollars) for both the motorized blind and daylight-harvesting lighting controls; we estimated that just the motorized blind portion of the system would cost $6.22 per ft² of glazing in 1998 dollars, and $12 per ft² of glazing in 2024 dollars.

So, per the numbers above, EC windows cost about 5x more than motorized blinds and would have a payback of well over 100 years in a retrofit scenario. It could be argued that a 29 year payback in the new construction / major renovation scenarios isn’t too far off the 10-year threshold for market acceptance, but the fact that spectrally selective low-e windows have a shorter estimated payback certainly weakens the value proposition for EC windows.

Undesirable Side-Effects

As previously noted, responsive daylight control’s potential to enhance the productivity, health, and wellness of building occupants is probably its most valuable benefit. However, responsive daylight control also has the potential to negatively impact the occupants if it operates in an erratic or annoying manner…and if that happens, there’s no way it could gain a foothold in the marketplace.

Unfortunately, responsive daylight control with current technology does, indeed, tend to operate in an erratic and annoying manner. We base this finding on our own testing of prototype responsive daylight control systems, as well as published data for a prominent automatic glare blocking system (whose issues are relevant to responsive daylight control as well as to automatic glare blocking).

We traced the propensity for erratic and annoying operation to two root causes:

• Conventional technology can’t reliably sense glare as people perceive it.
• Unlike the idealized curves of Figures 1 and 2, daylight irradiance doesn’t always vary smoothly over the daylight hours, but rather often fluctuates rapidly due to moving clouds.

Conventional Technology Can’t Reliably Sense Glare

People don’t sense daylight the same way a conventional light sensor does: they don’t just respond to irradiance, they also respond to contrast in the visual field.

As a result, it’s notoriously difficult to reliably estimate the perceived daylight level from the output of a conventional daylight sensor unless it happens to include a focal-plane array (i.e. a camera) that’s located and oriented to “see” the window from the same perspective as the building occupants. That’s obviously impractical in most situations.

And even such an optimally positioned sensor often won’t reliably sense glare caused by low-angle sunlight, especially in partly cloudy conditions. Despite having much lower irradiance than high-angle sunlight, low-angle sunlight can cause severe glare because it can penetrate deeply into a room.

This problem of reliably sensing glare applies to both automatic glare blocking and responsive daylight control systems.

Occupant Reaction to Automatic Glare Blocking in the New York Times Building

The most thoroughly documented daylight control installation of any type is in the New York Times Building in Manhattan, NY. Completed in 2007, the building includes an integrated shading-lighting system with automatic glare-blocking (based on roller shades) and daylight-harvesting capabilities.

LBNL performed a post-occupancy evaluation of occupancy satisfaction with the system (Clear, Robert D.; see References). This included a survey soliciting occupant feedback on issues related to visual comfort; this yielded 318 specific comments which LBNL broke out into 6 categories, as shown in the following chart:

The largest proportion (65%) of complaints were related to the automated shades, which were further broken out into 8 categories:

Fully 80% of the problems were due to the shades being in the wrong position. The report further states, “The most common concern with the window shades was that they failed to control glare. In addition, many employees felt that the shades operated in a meaningless (22) or inappropriate (28) manner, and were bothered by both too much glare when the shade was up, and too little light when the shade was down” (Ibid., page 14).

This was perhaps the most carefully specified and commissioned such system ever installed in an operating building and still represents the state-of-the-art in automatic glare blocking—and yet, per Figure 9, it wasn’t able to reliably sense or infer daylight glare as perceived by the occupants.

Two specific causes of dissatisfaction can be inferred from Figure 9:

• Some people felt the system was too aggressive controlling glare (“down when not needed”), while others felt the system wasn’t aggressive enough (“not down when needed”). It’s likely that this ambivalence was at least partially due to the fact that shade operation was synchronized by zone in order to maintain a more uniform external appearance. Because each zone included multiple occupants who saw the windows from different angular perspectives, a “one size fits all” synchronized shading setting would inevitably have been unsatisfying to some people.
• The number of “not down when needed” responses substantially exceeded the number of “down when not needed” responses. In other words, too much glare was a bigger problem than too little daylight. This implies that either the glare threshold for closing the shades was too high, or that the glare-sensing approach was unreliable under certain conditions.

Since glare must be avoided at all costs, Figure 9 implies that overall satisfaction could have been improved by making the system more aggressive in blocking glare, and in fact this was eventually implemented in the NY Times Building. However, this also necessarily lessened the value of the system relative to mostly-closed manually operated shading.

Daylight Irradiance Often Fluctuates Rapidly

The second cause of potentially annoying operation is a troublesome fact of nature: daylight irradiance often fluctuates wildly due to moving clouds. That’s shown in the following figure, which shows irradiance curves for typical clear-sky and moving-cloud conditions:

Note that daylight can, and often does, fluctuate more rapidly than the 6-minute averaging window of Figure 10 would suggest; 6-minute averaging was used for this plot because the data happened to be handy when writing this post.

The kind of fluctuation shown in Figure 10 isn’t an issue for a daylight control system using a Smart Window with no moving parts; such a system can adjust the shading as quickly and as frequently as the Smart Window technology will permit without any risk of annoying the building occupants. Unfortunately, as previously stated, Smart Windows are currently about five times more expensive than motorized window coverings—and will likely remain too expensive for mainstream use for the foreseeable future.

On the other hand, daylight fluctuation is a major issue for daylight control using a motorized window covering: no motorized window covering of any type could attempt to respond to the fluctuations of Figure 10 without thoroughly annoying the building occupants (and rapidly wearing out the motor).

There are two conventional approaches to dealing with this problem, both of which involve reducing responsiveness:

• Don’t attempt to respond to the fluctuation at all, but instead adjust the shading based solely on the movement of the sun (i.e. the orange clear-sky curve of Figure 10).
• Allow the system to respond to fluctuation, but only with enough inertia to avoid annoying the occupants. This can be done by low-pass filtering the sensor output, increasing the shade-actuation interval, or increasing the control dead-band (i.e. the allowable deviation from the desired daylight level before triggering a shading adjustment).

However, reducing responsiveness in either manner obviously also reduces the benefits of responsive daylight control relative to manually operated shading. Data from the previously-cited LBNL study puts this in perspective.

Impact of Deliberately Reduced Responsiveness

LBNL’s prototype automated blind system had a default activation interval of 30 seconds with unlimited blind movement per activation (Lee et al, Table 2, p. 23). A 30 second activation interval is 12 times shorter than the 6-minute averaging interval of the fluctuation shown in Figure 10, so the LBNL system would have attempted to respond to even more rapid fluctuation than shown in the figure.

Fortunately, their study involved only unoccupied spaces, so LBNL was able to investigate the impact of responsiveness on potential savings without having to worry about occupant annoyance. They observed that the savings decreased significantly when the activation interval was increased or the blind movement per activation was restricted:

• With an activation interval of 5 minutes or 15 minutes with limited angular movement per cycle, daily lighting energy use increased by 31-43% or 72-86%, respectively, on clear sunny summer days (Ibid., page 13).
• The 15 minute activation protocol increased daily lighting energy use by 24% on a partly cloudy summer day (Ibid.)

It’s not surprising that the benefit of responsive daylight control is reduced when responsiveness is reduced, but the amount of reduction for modest decreases in responsiveness might be surprising. The 5-minute activation interval investigated by LBNL is comparable to the 6-minute averaging interval used for the fluctuation curve of Figure 10, so even with this reduced level of responsiveness, the system would probably still have thoroughly annoyed occupants if they had been present.

Mitigating Risks of Annoying Operation Adversely Impacts Cost-Effectiveness

A variety of approaches have been proposed to mitigate the issues discussed above.

For example, there are two ways to increase the reliability of glare blocking:

• As previously mentioned, the glare threshold for closing the shading can be reduced—but this also reduces the average level of glare-free natural light, weakening the value proposition of automatic daylight control.
• A more sophisticated sensor (e.g. a camera co-located with the occupants) or multiple sensors can be used to sense glare more reliably—but this increases cost and complexity.

Similarly, there are straightforward approaches to mitigating potential occupant annoyance due to daylight fluctuation:

• As noted above, responsiveness can be deliberately reduced—but this also reduces the average level of glare-free natural light, weakening the value proposition of automatic daylight control.
• A Smart Window can be used instead of a motorized window covering, virtually eliminating the risks associated with daylight fluctuation—but only at far greater cost.

So, while there are recognized techniques to mitigate the risk of occupant annoyance, they either reduce the benefits or increase the cost of responsive daylight control. It might very well be possible to achieve enough cost-effectiveness and occupant acceptance for the mainstream market with current technologies, but that hasn’t happened yet.

Complexity

The two specific systems discussed above—LBNL’s prototype responsive daylight control system and the automated glare blocking system installed in the NY Times building—are significantly more complex than automated shading systems in general. They required remote sensors, computer-calculated solar trajectories, and specialized site-specific commissioning, and that’s true of any responsive daylight control or glare-blocking system based on current technology:

This level of complexity is ostensibly a significant disadvantage for a mainstream automated shading product, especially in the direct-to-consumer residential market.

An Untapped Opportunity That Could Finally Be Within Reach

Our market surveys show that responsive daylight control could indeed offer a compelling value proposition to mainstream non-residential and residential customers—if the benefits could be delivered at an acceptable price and without undesirable side-effects:

The maximum acceptable prices listed in Figure 12 are based on a nationwide average 10-year payback period for energy savings in office buildings, and on a market survey of potential customers for the residential market.

Obviously, Smart Windows have no hope of meeting these price thresholds in the near-term (if ever), so any mainstream solution is going to have to involve motorized window coverings—and the best choice, by far, is the motorized blind.

Back to the Future with Motorized Blinds

There were at least 5 good reasons why LBNL chose a motorized blind for their first responsive daylight control prototype back in the 1990s:

• The venetian blind is the only window covering that provides a continuously variable visible transmittance over the entire window area, like a Smart Window.
• Unlike other mechanical window coverings, a blind’s slat-tilt function is capable of fully blocking direct sunlight while still admitting useful substantial diffuse daylight—and not even a Smart Window can do that.
• A blind’s slat-tilt adjustments are much less obtrusive than raising/lowering a shade or opening/closing a curtain.
• The slat-tilt function is relatively easy and inexpensive to motorize.
• Over 650 million horizonal blinds are already in use in U.S. buildings (see the post linked in the next paragraph for how we obtained that number) and could be cost-effectively retrofitted for motorized operation.

These advantages are still valid today, and make motorized horizontal blinds the most cost-effective motorized window covering for almost any automated shading application—not just responsive daylight control. Indeed, motorized horizontal blinds are poised to eventually dominate the automated shading market in general.

Blinds Present a Unique Challenge of Their Own

Unfortunately, in addition to the previously described issues with motorized window coverings (i.e. potentially annoying operation, unreliable glare blocking, and complexity), motorized blinds present another unique challenge of their own: the variable shading provided by the slat-tilt function is direction-dependent. So, tilting the slats can decrease the daylight in one direction while increasing it another, greatly complicating the process of sensing or estimating the admitted daylight as it’s perceived by the building occupants:

• Since the shading provided by a blind is direction-dependent, a sensor won’t track the perceived daylight level as the blind is adjusted unless it “sees” the blind from the same perspective as the room occupants. Unfortunately, locating and orienting a sensor to achieve this is difficult in practice.
• In principle, this problem could be avoided by estimating, rather than directly sensing, the admitted daylight (e.g. from the output of a sensor on the outward-facing side of the blind and/or knowledge of the sun’s position relative to the window). In practice, however, the solar-optical properties of a blind are too complex to do this reliably.

Since glare is so deleterious, the previously-cited LBNL study restricted the slat tilt range of their automated blind to prevent a direct view of the sky, thereby avoiding the risk of glare from direct sunlight. However, they observed that this also significantly decreased the energy savings: when the slat tilt range was increased to allow a sky view, daily lighting energy use decreased by an average of 34% on partly cloudy to overcast spring days, and daily cooling loads decreased by an average of 5% (Lee et al., p. 13).

So, while blinds are probably the best bet for cost-effective responsive daylight control in the near-term, there are still significant challenges in using them for that purpose.

Can the Challenges be Overcome?

LBNL didn’t attempt to further refine their blind-based prototypes after their full-scale testing in unoccupied offices in the late 1990s; instead, they shifted their focus to Smart Windows—and manufacturers who might have been interested in responsive daylight control seem to have followed LBNL’s lead.

That’s too bad, because it’s conceivable that some additional testing and tweaking might be enough to turn a system like LBNL’s prototype into a mainstream product.

Also, several new technologies have emerged since the late 90’s that could be relevant to a system like LBNL’s prototype:

• Inexpensive, high-performance digital light-sensing chips (driven by the proliferation of cameras in personal electronic devices) could enable more effective sensing of daylight glare.
• Inexpensive microcontrollers with wireless connectivity and interoperability-oriented networking standards like Matter could facilitate integration between shading, sensing, and lighting hardware and cloud-based sources of weather information.
• Wider use of building-automation platforms in both non-residential and residential buildings could enable more sophisticated sensor processing and control algorithms.

These technologies are being leveraged by automated shading products in general, but they seemingly yet haven’t been applied to meet the unique challenges of responsive daylight control.

References

Clear, Robert D. “Post-Occupancy Evaluation of The New York Times Headquarters Building: an Examination of Causes for Occupant Satisfaction and Dissatisfaction with the Energy-Efficiency Measures”. Ernest Orlando Lawrence Berkeley National Laboratory, technical memorandum. 2010. <https://facades.lbl.gov/motorized-shades>

Energy Information Administration, Commercial Buildings Energy Consumption Survey. “2018 CBECS Survey Data”. <https://www.eia.gov/consumption/commercial/data/2018/#b11-b14>

General Services Administration, “ELECTROCHROMIC WINDOWS FOR OFFICE SPACE”. Green Proving Ground assessment GPG-033. November 2017. <https://www.gsa.gov/system/files/033-Findings-EC_Windows_for_Office_Space-v1.pdf>

Heschong Mahone Group, Inc. “Daylight and Retail Sales.” 2003. <https://newbuildings.org/wp-content/uploads/2015/11/A-5_Daylgt_Retail_2.3.71.pdf>

Heschong Mahone Group, Inc. “Daylighting in Schools: An Investigation into the Relationship between Day lighting and Human Performance.” July 21, 1999. <https://eric.ed.gov/?id=ED444337>

Heschong Mahone Group, Inc. “Sidelighting Photocontrols Field Study”. Report prepared for the Northwest Energy Efficiency Alliance, Pacific Gas and Electric Company, and Southern California Edison Company. 2005. <Sidelighting-Photocontrols-Field-Study>

Heschong Mahone Group, Inc. “Skylighting and Retail Sales: An Investigation into the Relationship Between Daylighting and Human Performance.” 1999. <https://www.researchgate.net/publication/328965592_Skylighting_and_Retail_Sales_An_Investigation_into_the_Relationship_Between_Daylighting_and_Human_Performance_Condensed_Report_for_PGE_by_the_Heschong_Mahone_Group_1999>

Heschong Mahone Group, Inc. “Windows and Offices: A Study of Office Worker Performance and the Indoor Environment.” October 2003. <https://newbuildings.org/wp-content/uploads/2015/11/A-9_Windows_Offices_2.6.101.pdf>

Lee, E. S., D.L. DiBartolomeo, E.L. Vine, and S.E. Selkowitz. “Integrated Performance of an Automated Venetian Blind/Electric Lighting System in a Full-Scale Private Office.” Ernest Orlando Lawrence Berkeley National Laboratory, Report LBNL-41443, 1998.

The Lighting Research Center, Rensselaer Polytechnic Institute. “Daylight Dividends”. Webpage, publication date unknown. <https://www.lrc.rpi.edu/programs/daylighting/dr_productivity.asp>

The Velux Group. “1.4 Benefits of daylight.” Webpage, publication date unknown. <https://www.velux.com/what-we-do/research-and-knowledge/deic-basic-book/daylight/benefits-of-daylight>