In January 2012, Katrin Klingenberg, the founder of the Passive House Institute U.S. (PHIUS), announced that her organization would develop a new passive house standard for North America — a standard that differed from the Passivhaus standard developed in Darmstadt, Germany.
Writing in her blog in 2012, Klingenberg explained that “it’s time to allow for a modification process to the rigid annual heating and cooling requirement of less or equal to 15 kWh/m²•yr… for the North American continent’s more extreme climates… This idea that we need to adapt the standard to various regions has taken root around the world from domestic energy experts like Martin Holladay, Alex Wilson, and Marc Rosenbaum and to Passive House groups from other countries, like the Swedes.”
Almost two years later, Klingenberg made another announcement: the work required to develop the new standard would be partly funded by U.S. taxpayers (through the Department of Energy), and one of the contracts for the required study would be awarded to the Building Science Corporation of Westford, Massachusetts.
A few weeks ago, Klingenberg’s first goal was reached when the DOE-funded paper (“Climate-Specific Passive Building Standards”) was published. The report has three authors: Betsy Pettit (from Building Science Corporation), Graham Wright (from PHIUS), and Katrin Klingenberg.
In effect, the paper is a draft for a proposed new passive house standard. PHIUS is now inviting the public to comment on the draft standard; after the public comments are reviewed, the standard may be modified before being adopted by PHIUS.
What is a passive house?
A few years ago, PHIUS cut the umbilical cord linking it to Germany. Since then, the U.S. organization has no longer been bound by the German definition of a “Passivhaus.” Because if its recent independence, PHIUS now has a chance to ask an important question: what, exactly, do we mean by a “passive house”?
Back in the late 1970s, the term “passive” was used to differentiate two different approaches to solar architecture. The “active solar” approach included either a system with fluid-filled solar collectors and a pump that circulated fluid from a storage tank to the collectors, or a system with a fan that circulated air from a bin filled with rocks to roof-mounted solar-air collectors.
The “passive solar” approach did not include any pumps or fans. Instead, it depended on oversized south-facing windows, interior thermal mass, and properly sized roof overhangs.
While the latest PHIUS document hints at this early distinction between the “passive” and “active” approaches — the new paper says that “PHIUS acknowledges that passive house was born in Canada and the U.S. in name and concept” — it has come up with a new definition of the “passive” approach. The draft document states: “The view of what constitutes a passive measure … includes fan- and pump-assisted devices such as HRVs, earth air tubes, brine loops, and whole-house fans, in addition to insulation, air-sealing, overhangs and such.”
For designers who are old enough to remember the 1970s, such a definition is startling. When I saw this ahistorical definition, I felt like putting my forehead on the table. But by now, any attempt to reclaim the old definition of a “passive solar house” seems doomed to failure. Evidently, the muddying of the waters that began when Dr. Wolfgang Feist established a definition for “Passivhaus” in the 1990s can’t be fixed. It’s now too late to untangle the string.
The new PHIUS document notes, “As passive house standard adaptations go, the one described here is relatively far-reaching. Nevertheless it retains all defining characteristics of a ‘passive’ building.” If only the confused reader knew what a “passive building” is, perhaps these two sentences might make sense.
For the time being, however — at least until the new standard is finalized — we’ll all have to shake our heads in dismay and avoid asking the embarrassing question of what anyone means these days when they use the phrase “passive house.”
What is wrong with the existing Passivhaus standard?
The new PHIUS document alludes to at least 14 problems with the existing Passivhaus standard developed in Germany.
1. A single rigid standard — one requiring a maximum peak heating load of 10 W/m² or a maximum annual heating energy budget of 4.75 kBtu/ft²•yr — can’t possibly represent the “economic optimum” approach for every location on the planet. As the draft proposal notes, “PHI [the Passivhaus Institut in Germany] claims that the ‘economic optimum’ occurs at 10 W/m² peak heat load or the 4.75 kBtu/ft²•yr annual heat demand everywhere in the world. ‘That can’t be right’ is the objection.”
2. North American climates aren’t comparable to European climates. While the number of heating degree days (which influence the annual heating fuel budget required for a building) for many locations in North America is similar to the number of heating degree days in European locations, North American locations often have higher design heating loads (which influence the size of the needed heating system). As the PHIUS paper notes, “the relation between degree-days and peak design temperature varies by climate; they are but weakly correlated. Away from the [Atlantic and Pacific] coasts [in North America], peak design conditions are relatively harsh compared to degree-days.”
The paper also notes, “PHI literature usually quotes -10°C/14°F as a peak load design temperature for central Europe … In the mid-continental United States, places with similar heating degree-days to Germany have much harsher design temperatures.”
The fact that many North American locations have colder design heating loads than German locations means that our thermal envelopes have to include much more insulation to meet the 15 kWh/m²•year (4.75 kBtu/ft²•yr) heating energy budget than would be the case in Germany. The insulation required to meet the German energy budget are so thick that they aren’t cost-effective.
3. “Tunneling through the cost barrier” didn’t work. Wolfgang Feist, the German physicist who developed the Passivhaus standard, claimed that Passivhaus levels of airtightness and insulation save money, because a Passivhaus building doesn’t need a conventional heating system. This idea — that it is possible to add so much insulation to a building that savings can occur by simplifying the building’s HVAC system — was termed “tunneling through the cost barrier” by Amory Lovins.
The “tunneling through the cost barrier” idea is like a unicorn: often described but rarely seen. As the PHIUS report notes, “In North America, ‘tunneling through the cost barrier’ was not achieved. Unlike Germany, there is not such a clear breakpoint where an expensive baseline boiler and hydronic distribution system (the typical heating system in Europe) can be eliminated for great savings.”
Compared to conventional buildings, most Passivhaus projects in North America have seen significant cost premiums rather than cost savings. The PHIUS report notes, “In a 2009 article, John Straube critiqued PHI’s standard. While this article contained some misunderstandings, its basic point was accurate that in ASHRAE Climate Zones 5 through 7 in North America, the standard is not economically justifiable, by and large.”
4. When used in North America, the existing Passivhaus standard results in overglazed buildings that sometimes overheat. Several GBA articles, include one that I wrote in 2012, have noted that many Passivhaus buildings in the U.S. have too much south-facing glazing. The new report acknowledges that this is indeed a problem: “PHIUS+ certification that uses European energy metrics and specific standards as written has resulted in (broadly speaking) passive-solar-esque designs with a tendency to overheating, and discouragingly high cost premiums.”
5. The existing standard’s default values for miscellaneous electrical loads are unrealistically low. PHIUS would fix this problem by setting new (more realistic) default values for electrical loads.
6. The existing standard’s default value for domestic hot water use is unrealistically low. PHIUS would fix this problem by setting a new (more realistic) default value for domestic hot water use.
7. The European method of calculating a building’s floor area is quirky and cumbersome. The PHPP spreadsheet that determines whether a proposed design gets a “ja” or “nein” has an unusual definition of floor area. The definition is taken from a German law, the WohnflÃ¤chenverordnung, that defines floor area as the interior area of a building, excluding the following elements: the exterior walls, the plaster on walls, the area of chimneys, the area of interior partitions, the area defined by an interior door, the area of columns, and the area taken up by stairs with more than three steps. Other caveats: rooms with a ceiling height below 6.6 feet are excluded (although rooms with ceiling heights between 3.3 feet and 6.6 feet in height are added at 50 percent of the area); the rooms must be within the thermal envelope (so that three-season rooms and garages are excluded); and storage and mechanical rooms inside the thermal envelope are included at 60 percent of the area.
It’s safe to say that no one in North America (other than Passivhaus consultants) defines a building’s floor area following the WohnflÃ¤chenverordnung rules. The new draft document from PHIUS proposes a new, easier-to-calculate method for determining a building’s floor area.
8. The existing Passivhaus standard discourages the use of North American HRVs and ERVs by imposing an unjustifiably harsh method for de-rating published efficiency ratings for these North American appliances. To correct this problem, the draft standard proposes a new formula for calculating the efficiency of HRVs and ERVs.
9. The existing Passivhaus standard imposes a source energy factor for grid electricity that is based on the European grid, not the North American grid. The draft standard proposes a new, more realistic source energy factor for North America.
10. The existing Passivhaus standard bases its source energy limit on a building’s area rather than on the number of occupants. The draft standard notes that a per-person limit makes more sense.
11. The existing Passivhaus standard allows the energy produced by an on-site solar thermal system, but not the energy produced by an on-site PV system, to “count” toward meeting the standard’s energy targets. The draft standard will correct this problem by allowing some of the energy produced by an on-site PV system to be counted.
12. The PHPP spreadsheet assumes a high (and possibly erroneous) rate of heat loss in winter through concrete slabs on grade. The PHPP calculations for heat loss through a slab on grade differ from similar calculations made by North American energy modeling programs. Because of this discrepancy, the PHPP software ends up validating building designs that include very thick layers of sub-slab rigid foam — foam that is so thick that the approach leaves North American observers scratching their heads. Here’s how the problem is described in the PHIUS paper: “[The] Ground contact calculation protocol [is] very different between EnergyPlus dynamic and PHPP/WP static. Anecdotal evidence suggests that the EnergPlus’ method predicts a lot less heat loss to the ground than ISO 13370-based static calculations. If so and if EnergyPlus is right, then designers using PHPP/WP are over-insulating their floors.” Although the issue was raised by the technical committee, no conclusions were reached. The paper simply notes, “This discrepancy needs to be confirmed and corrected.”
13. The existing Passivhaus standard is based on modeled performance, not measured performance. While hinting that this problem exists, the committee that drafted the new proposed standard decided not to address it. The draft paper notes, “The proposed adapted standard is still performance-based, that is, based mostly on modeled performance, as opposed to a prescriptive approach or an outcome-based approach.” One proposed solution to this problem would be to delay awarding Passive House certification until homeowners could show one year of energy monitoring data that matched or beat the modeled projections; the committee did not adopt this suggestion.
14. The existing Passivhaus standard includes a small-house penalty. Because larger homes have a smaller surface-to-volume ratio than smaller homes, it’s easier for a large home than a small home to meet the Passivhaus standard. The committee that drafted the new standard alluded to the problem but decided not to fix it: “The studies [made by the PHIUS committee] are predicated on providing housing that is typical for the North America market (i.e. the three-bedroom house). More efficient forms of housing, such as multifamily units, will have an easier time meeting the criteria, while less efficient forms, such as detached ‘tiny houses’ will have a harder time of it.”
Did the committee aim for cost-effectiveness?
The authors of the report refer to themselves as the technical committee (TC for short). While many people assumed that the committee would pay attention to cost effectiveness, in fact the authors left themselves a little wiggle room. Instead of creating new standards that are cost-effective, the committee decided to aim for standards they call “cost-competitive.” Here’s what the authors wrote: “The reported work is an up-to-date, independent study of how much investment in passive measures can be economically justified as cost-competitive, if not strictly cost-optimal.”
The main tool used for this determination was BEopt software. (For more information on BEopt, see BEopt Software Has Been Released to the Public.) BEopt software aims to optimize the specifications for a building envelope, so that the cost to the owner of a new home for utility bills plus the mortgage are as low as possible (see Image #2, below). The software recognizes that beyond-code improvements to a home’s envelope can result in lower total costs, even when the mortgage is larger than it would be for a code-minimum home — but only up to a point (the cost-optimum point).
The PHIUS committee decided that there were reasons to pay for additional envelope improvements, even if these envelope improvements go beyond the cost-optimum point and result in higher total costs for the homeowner.
Here’s the approach that the committee took: “A human judgment call was made, as to the point of deepest energy savings feasible, cost-competitively — location by location. … The PV start point would be a defensible level at which to set the criteria. But it may be appropriate to choose a more aggressive point on the cost-optimal curve, that is, one still cost-competitive but with less annual dollar savings. …The rationale is that passive measures are better for the building owners and occupants than renewable generation alone. They increase the building’s resilience to utility outages, by minimizing heat losses and thus allowing interior temperature ‘coasting’ during outages. … The TC [technical committee] as a whole was inclined to forgo some annual dollar savings if more peak load reductions could be realized. … One could argue that pushing past the cost optimum is actually a conservative approach given the uncertainty of … future developments and possible climate risks. … The TC agreed upon the following heuristic for setting the criteria: … Note the knee of conservation-only cost curve and go a little past it, to where conservation is heading into diminishing returns.”
In other words, expensive envelope measures are sometimes defensible, even when the cost of these measures is somewhat higher than can be justified by the expected energy savings.
A modeling exercise
The process described above — using BEopt software to determine which beyond-code envelope improvements result in a house with the lowest total cost for the homeowner — was applied to a “typical” two-story, three-bedroom house. Once this cost-optimum point was determined, the committee used human judgment to decide what additional envelope measures to include beyond the cost-optimum point.
The modeled house was an all-electric house with a slab-on-grade foundation, a vented attic with cellulose on the attic floor, and walls insulated on the exterior with a continuous layer of EPS foam. The technical committee decided to include a “constraint,” namely that windows had to have a relatively low U-factor — low enough to ensure that the interior surface of the glass stayed at 60°F or warmer during the winter. The committee wisely decided that “there will be no subsidizing performance upgrades by cheapening finishes. This strategy, while effective if you can get it on a project, is unfair to include in the studies.”
The energy performance of this house was modeled in over 100 North American locations (see Image #1 at the top of the page).
The committee’s recommendations
The fundamental problem that the committee set out to address is that the existing Passivhaus standard, when implemented in North America, often results in buildings that aren’t cost-effective. These Passivhaus buildings have insulation that is so thick, and windows that are so expensive, that the cost of the insulation and window upgrades is much higher than the value of the energy that will be saved by these upgrades over the life of the building.
In response to this fundamental concern, the committee proposed different energy use targets for space conditioning and different window U-factor specifications for a variety of climates. These proposed targets and specs are shown in Table 8 of the committee’s report (see Image #3, below).
While the existing Passivhaus standard allows designers to aim for either a maximum annual energy budget for space conditioning of 15 kWh/m²•year or a maximum peak heating load of 10 watts/m², the proposed new standard will require that both targets — a target for the annual energy budget for space conditioning and the target for maximum peak heating and cooling loads — be met. As the report notes, “We propose to set limits on annual heat demand and peak heating load, as well as annual cooling demand and peak cooling load. So the criteria would read: Annual heating demand < A, and Annual cooling demand (sensible+latent) < B, and Peak heating load < C, and Peak cooling load < D. These would vary by climate.” Instead of the current Passivhaus space heating energy budget of 15 kWh/m²•year (equivalent to 4.75 kBtu/ft²•yr), the committee proposes less rigorous targets for Climate Zones 5 through 8 and for some locations in Climate Zone 4. These new proposed targets would range from 5.6 kBtu/ft²•yr in Climate Zone 5B to 13.2 kBtu/ft²•yr in Climate Zone 8. The committee proposed more rigorous space heating budgets for Climate Zones 1 through 3, ranging from 0 kBtu/ft²•yr in Climate Zone 1A to 3.0 kBtu/ft²•yr in Climate Zone 3A.
Cooling energy budgets would also be adjusted, and would range from 0.2 kBtu/ft²•yr in Climate Zone 8 to 18.6 kBtu/ft²•yr in Climate Zone 1A.
These proposed space heating and cooling energy budgets are not set in stone. In fact, the paper notes, “the TC [technical committee] doesn’t think a tabular approach like this [that is, an approach like the one presented in its report] is granular enough for program use.” Presumably, a more refined method will be proposed before any new Passive House standard is approved by PHIUS.
In addition to these proposed space heating and cooling energy budgets — targets that cannot be exceeded — the committee proposed climate-specific peak heating load targets and peak cooling load targets.
The peak heating load targets range from 1.75 Btuh/ft² in Climate Zone 1A to 8.4 Btuh/ft² in Climate Zone 8.
The peak cooling load targets range from 10.7 Btuh/ft² in Climate Zone 2B to 4.9 Btuh/ft² in Climate Zone 3C.
The recommended maximum glazing U-factor for windows ranges from U-0.40 in Climate Zone 3C to U-0.10 in Climate Zone 8. These proposed maximum U-factors are driven by occupant comfort concerns, not energy budget concerns.
In addition to the proposals listed above, the committee made several other recommendations.
A new way to calculate floor area. Here is the new proposed definition: floor area will be “measured on the interior dimensions of the passive house thermal envelope, drywall-to-drywall, where ceiling height is greater than or equal to seven feet. This specifically includes stairs and interior partitions, as well as baseboards and cabinets. It specifically excludes open-to-below.”
A new way to calculate the efficiency of HRVs and ERVs. The committee has set itself the goal of making this change, but a proposed solution has not yet been published. The report notes that “the efficiency ratings of heat-recovery ventilators aren’t apples-to-apples comparable between PHI and domestic institutes (HVI and AHRI). Up to now PHIUS has been using a rule of thumb from PHI, ‘subtract 12% from the sensible efficiency of non-PHI-rated units.’ The TC [technical committee] recently determined more nuanced adjustments to HVI and AHRI ratings that bring them closer to comparability with PHI rating, and the 12% deduction remains only for units that don’t have any third-party rating. This work is also beyond the scope of this report and is being written up separately.”
A new default value for domestic hot water use. The proposed new standard assumes that occupants “Use hot water as per BA [Building America] assumptions (~50% higher than PHPP).”
A new default value for miscellaneous electrical loads. The committee recommended that “For residential projects, the defaults for lighting and plug loads [should] increase to 80% of RESNET levels.” This is a big change: “These [new default levels] are about six times the PHPP defaults but lower than Building America baseline home. … The low PHPP defaults are grossly unrealistic, a discrepancy that must be fixed.”
This change will have several interesting consequences. “Such an increase in residential lighting and plug load defaults is a large change that makes it considerably harder to meet the source energy target. … Under the PHI protocol [that is, the existing Passivhaus standard], the space conditioning criteria were usually the limiting factor, while the source energy target was relatively easy to meet. But with higher lighting and plug load defaults, and potentially higher space conditioning thresholds, the source energy limit could become the limiting factor.”
A new source energy cap. The source energy cap is the maximum amount of source energy that a Passivhaus project can use. (For more information on the distinction between source energy and site energy, see Understanding Energy Units.) In the existing Passivhaus standard, the source energy cap (120 kWh/m²•year) is based on a building’s area. PHIUS proposes a change: “For residential projects it is appropriate to change [the source energy cap] to a per-person budget, based on a fair-share of the atmosphere consideration. Occupancy is therefore taken to be the number of bedrooms plus one, per dwelling unit.” Elsewhere, the report notes, “Straightforward conversion of the 120 kWh/m²•year limit times 35 m²/person standard occupancy would give a limit of 4,200 kWh/person•yr. A review of previously certified [Passivhaus] projects showed a median source energy design for 4,100 kWh/person•yr, but with lighting and plug load defaults adjusted to RESNET levels, the median would have been almost 6,600 kWh/person•yr.”
In other words, the new proposed source energy cap is so stringent that many certified Passivhaus buildings would be unable to meet it. The committee therefore came up with a compromise solution, proposing that “as a shock absorber, the source energy limit should be temporarily relieved [changed] to 6,000 kWh/person•yr, returning [rather, changing again] to 4,200 by a date to be determined.”
A new source energy factor for grid electricity. The committee recommends changing this energy factor from 2.7 to 3.1. The report notes, “The U.S. electric grid is known to have source energy factors ranging from 2.374 to 3.549 depending on the major interconnect region, with a national average of 3.138. For the sake of simplicity and a level playing field, it is reasonable to use the national average. In recognition that the grid has probably gotten cleaner since the report was published, one can round down.”
A change allowing electricity produced by PV arrays to “count” toward Passive House targets. The committee decided to allow PV electricity produced on site to “count,” but only if the PV electricity is used on site simultaneously. The report notes, “Currently, the only renewable energy that ‘counts’ towards reducing source energy is solar thermal. The Committee agreed to put other renewable generation on the same footing if it is used as it is produced.”
New airtightness limits. Instead of setting an airtightness limit at 0.6 air changes per hour at a pressure difference of 50 pascals — the existing target — the committee proposed a new limit using different units. The proposed limit would be 0.05 cfm50 per square foot of gross envelope area. As the paper notes, this proposed change “allows the airtightness requirement to scale appropriately based on building size. Before, a larger building that met the 0.6 ach50 requirement could be in actuality up to seven times more leaky than a small single-family home that tested the same.”
More on airtightness
The PHIUS paper repeats a problematic claim made by Wolfgang Feist — namely, that “The airtightness requirement comes from consideration of building durability and mold risk.”
No convincing argument has ever been presented to show that the 0.6 ach50 target is necessary to prevent condensation, mold, or rot. On the contrary, there is plenty of evidence that buildings with air leakage rates of 0.6 to 2.0 ach50 are performing very well.
Instead of abandoning Feist’s basis for establishing an airtightness target, the PHIUS committee appears to have embraced it — while simultaneously noting that the target will need to be changed if the argument for its basis is retained.
The paper notes, “The top priorities for future work at this point are: … Studies on relaxing the airtightness criteria by climate. Again, the airtightness requirement is driven by moisture risk (energy savings being a side benefit). It stands to reason that the danger threshold would be climate-dependent. Also, it may be appropriate to revisit the field testing protocol: perhaps the test should be done two different ways — one for energy modeling purposes being realistic about leakage in normal operation, and another protocol for durability, focusing on leakage through the assemblies, with more of the nonthreatening things like door thresholds and vent dampers taped off. … The airtightness requirement was reconsidered on the basis of avoiding moisture and mold risk, using dynamic hygrothermal simulations to be published elsewhere.”
It’s all about electricity rates
As an aside, the paper includes two key sentences that are startling: “In places with expensive energy, everything was affordable in a sense: even [envelope] measures that were deep into diminishing returns still showed cash flow. In places with cheap energy, distressingly little was affordable.”
The implications of these two sentences are profound. If the PHIUS committee expects its report to be taken seriously — and clearly it does — then readers are owed a much deeper exploration of the implications of these two sentences than the report provides.
It’s a fact that upgraded envelope measures aren’t cost-effective in areas with cheap electricity. Anyone who is developing a green building rating program needs to address this issue head-on. The members of the PHIUS committee are to be commended for tackling the thorny problem of cost-effectiveness, but their efforts have led them down an alley to a place where they didn’t want to end up.
The PHIUS paper suggests several reasons (including increased resiliency) to defend envelope measures that aren’t cost-effective. But the committee hasn’t clearly addressed the fact that areas with cheap electricity don’t really need fancy thermal envelopes. This fact becomes especially stark during historical periods (like this one) when energy prices (including prices for PV power) are dropping.
No one knows what a “passive house” is
This dilemma points to an unresolved question: what, exactly, is a “passive house”? If PHIUS believes that a passive house can include a furnace, a ductless minisplit, an active solar thermal system, an HRV, and a buried brine loop hooked up to a circulating pump, it appears as if we have entered an Orwellian world where “passive” means “active.”
This dilemma is especially poignant for PHIUS, since PHIUS wants to distinguish the “passive house” approach from the “net zero energy” approach. But unlike the members of the PHIUS committee, who have yet to come up with a cogent definition of “passive house,” at least the members of the net-zero-energy community know how to define a net-zero-energy house. For the net-zero-energy designer, an envelope measure makes sense if the value of the energy saved by the measure exceeds the value of the energy saved by a PV system that costs as much as the envelope measure under consideration.
If the PHIUS committee adopted this approach, at least there would be a rational way to defend upgraded envelope measures in regions where energy is so cheap that “distressingly little is affordable.” The only remaining problem for PHIUS (if they took this advice) is that there would no longer be any difference between the net-zero-energy approach and the “passive house” approach.
The committee deserves praise
Katrin Klingenberg and the other members of the PHIUS technical committee have embarked on a very interesting journey. Their willingness to listen to technical arguments undermining the foundations of the Passivhaus standard has been remarkable, especially in light of the fact that most of the committee members have been proponents of the Passivhaus approach for many years. The committee’s proposed changes represent useful and defensible improvements to the Passivhaus standard.
It will be interesting to see whether the new PHIUS construction standards gain traction in the U.S. over the coming years, or whether the PHIUS approach will remain (as it is now) a little-used certification program that is difficult to implement without the help of a consultant trained in the use of a complicated spreadsheet with a daunting number of inputs and calculations.
Is the concept compelling enough to attract homeowners?
If homeowners choose to follow the PHIUS path, what kind of a house will they end up with?
It won’t be passive. Like most homes in the U.S., it will require active HVAC equipment to supply space heating, cooling, ventilation, and domestic hot water.
It won’t have the lowest possible ownership cost (that is, mortgage cost plus utility costs), because the PHIUS committee has decided to require investments in envelope measures that push the cost of a new home beyond the cost-optimum point.
It probably won’t be a net-zero-energy house, because the PHIUS committee isn’t considering a requirement for a PV array that can balance annual energy use.
Time will tell whether there is much of a market for this type of home in the U.S.
Martin Holladay’s previous blog: “Solar Thermal Is Really, Really Dead.”