Internship: Demographic meta-analysis of the Sierra College student body and disparately impacted success of its populations.

Executive Summary

During the fall 2019 semester, I completed a sociology internship at Sierra College. At some point in the last year, I had asked several professors which populations are the most impacted at Sierra College (Which group has the lowest success). I was surprised to learn that no one knew the answer. Further research led me to learn that there was no current way of searching the data for the answer to this question. I set out to create a list of all populations which could be filtered and sorted by success.

The school is responsible for improving conditions for impacted populations. Sierra College’s current approach to closing equity gaps is that three committees work separately on an arbitrary set of less than half of the demographics for which we are reporting success data. None of these committees focuses on intersectional identities. None of these committees is funded. None of these committees has any real power outside of making recommendations.

For nearly all of the most impacted groups, their identities are not addressed by this arbitrary structure, and their success metrics are getting worse, not better.

In many cases, the current system seems designed to hide the problems or make it difficult to see them. The college works hard to exclude third party experts from critical processes and to suppress important data related to the success of impacted populations.

What Is Disparate Impact?

Disparate impact means that people in some groups do not succeed to the same degree as people in other groups do, on average. Equity gaps measure how big the difference is between the lower success of some groups versus the higher success of some other group. In sociology, this is a major field of research. California recently passed new rules which base funding for schools on how well they close these gaps over time.

In sociology, we track several major impacted populations. We know for example that marginalized classes (such as people of color, women, transgender people, and others) will have disparate impact because these groups are suffering marginalization in society. Black people will be impacted versus white people. Women will be impacted versus men. Typically these differences are measured in terms of life expectancy, income, housing stability, etc. In this case we are measuring success at Sierra College for each group.

Who Is Most Impacted?

At some point in the last year, I asked several professors which populations are the most impacted at Sierra College. I was surprised to learn that no one knew the answer. Further research led me to learn that there was no current way of searching the data for the answer to this question. One could, for example, check the success of “gay black transgender students,” but there was no list of all the populations which could be sorted by their success metrics.

During the fall 2019 semester, I completed a sociology internship at Sierra College. The project was to build a new way of analyzing disparate impact. Over the course of the semester, I created a new database which contains all the public success data for all the groups which are large enough to be reported. A law called FERPA says that we can not report the success of populations with a size smaller than ten.

Initially, I tried to execute this analysis through the school’s research department. FERPA became a favorite roadblock which would be raised when I asked questions or asked for data. “[We can’t answer that because the answer could include data about groups with population sizes fewer than ten, so it would violate FERPA.]” When I followed up to suggest that we simply include the population size in the results, and remove any results where the size is less than ten, the response was that this would be infeasible. I quickly decided to simply conduct the meta-analysis myself.

Results and Methodology

My final report is available here. Demographics and population size are listed for all the populations, while success metrics are listed to the right. The data can be sorted by any value, and filters may be used in the search box. Additionally, I have also published the database as well as the tools I built in order to complete this project. Free tools such as SQLite Browser will allow anyone to query and analyze the database further.

The schools’ current equity strategy looks at identities from a non-intersectional  perspective. Sierra College has three committees working on student equity. One committee focuses on race, one committee focuses on gender, and one focuses on sexual orientation as well as gender-identity. None of these committees is funded or has any real power except to make recommendations. I want to repeat this point because it is the critical finding of this research. Sierra College’s current equity committees look at gender OR race OR sexual orientation/transgender. Sierra College’s current equity committees DO NOT look at gender AND race or gender AND sexual orientation. None of the equity committees look at the many other demographics such as foster youth status, income, veteran status, disabilities, homelessness, etc.

The system of looking only at one dimension of identity at a time while not comparing them to others or including multiple dimensions leaves huge blind spots for the people working on closing equity gaps. I see two main problems that come from this  blind spot;

  1. The first major problem arising from this blind spot is that we don’t know what we don’t know. For example, all ten of the most impacted populations are black, but eight of them are also former foster youth. This highly significant factor is not addressed by any of the current equity committees. There are programs for foster youth, and there are programs for black students, but there is no one whose job it is to look at improving conditions for black foster youth. It’s not surprising that for all ten of these populations, conditions are getting worse, not better.
  2. The second major problem is the lack of an intersectional approach to closing equity gaps. We pay a lot of attention to race, but the fact that the most significant factor for black students is an intersection with foster youth status was previously unknown. It’s worth mentioning that none of the equity committees looks at foster youth status. In fact, while the school reports eight dimensions of demographics for success data, the equity committees are looking at only four of them. Several of the state’s required metrics are not being reported at all. (More on that later.) Despite the fact that foster youth status is the most significant factor impacting success of black students, no one is empowered to work on the issues affecting specifically black foster youth students.

This is not just about black foster youth students. Throughout the meta-analysis, we see examples where complex intersectional identities show disparate impact which one-dimensional identities do not. We see hundreds of students whose intersectional identities are not addressed or even observed by the current one-dimensional system. For example;

  • The ten most impacted LGBT groups are all non-white
  • The ten most impacted female groups are all non-white
  • The ten most impacted low-income groups are 70% non-white

In all of these examples, almost all of the intersectional groups show success going down, while success for the corresponding one-dimensional identities is going up. This is a critical flaw in the current system. When white women and black women are categorized together, the average looks like women are doing ok with a 0% change this year. But in reality, black women are seeing low and decreasing success while white women are seeing high and increasing success. The average is inaccurate and ignores the fact that people have intersectional identities.

  • Success of multi-ethnic homosexuals (n=23) went down by 24% this year.
    • Success of homosexuals alone (n=249) only dropped 4%, buoyed by white homosexuals (n=131) which saw a 3% increase in success, while some white subgroups saw success increases as high as 29%.
  • Success of low-income black women who were foster youth went down by 31% this year.
    • Females overall (n=12,594) saw a 0% change in success this year, buoyed by more privileged groups.

My interviews with the equity committee members have highlighted another problem. Since these committees are not funded, they have to spend most of their time fundraising, and then typically, some time each year is dedicated to throwing an event to highlight the problems facing their communities. At no point in the year are these committees actually working directly on improving success. They are focused on community engagement, a prerequisite for success. If these committees were funded, they could focus on the actual work which they ostensibly exist to do, instead of focusing on fundraising. One idea that came up in interviews was that the committees could create proposals for projects or programs which would work to increase success. Then they could be funded by the district to execute these projects, since this is the basis of the district’s funding under the new funding formula.

The value of third-party experts can not be exaggerated in this work. One student organization, “Salaam Sierra” is lucky enough to have a chaplain who comes in to work with them. This person is aware of resources at Sierra and throughout the broader community, they retain institutional memory as members of the organization come and go, and they can advocate for the organization on campus and in the wider community. All student organizations, academic committees, and other groups focused on equity and success should have expert third party representatives to facilitate their work.

In some cases, the college has actively worked to exclude expert third parties from assisting student organizations, in direct violation of the law. For example, one administrator said in a recent Q&A panel that when staff hear that students have been sexually assaulted, they are forbidden from following the law and reporting directly to the police. Instead, she said, the staff are directed to report sexual assaults to a bureaucrat at the college who will decide whether the sexual assault should be reported to the police. This is a clear violation of the mandated reporter law, and of Title IX.  Expert third parties in the community such as the local rape crisis agency even offered to pay for trained crisis counselors to be present on campus and accessible to students who need them. Instead, the administrator in question insists that untrained academic counselors are more than adequate for recording the complaints of students and forwarding them to bureaucrats rather than following the law and reporting to the police. This is just one of many examples where any success of students at Sierra College is  in spite of the institution and not because of it. Allowing competent third party experts to be present would no doubt bring problems like sexual assault on campus to light and negatively impact the image of the college, and therefore the corrupt leadership insists on excluding the experts who could help improve the success of students on campus.

We need independent third party experts to be invited and welcomed and included to participate in engagement centers and student organizations as well as academic senate committees and especially equity committees. Imagine that clubs for example must have not just a faculty adviser but also a community adviser; suddenly the black student union has a club structure which requires them to seek out a community activist or leader who can give them far more resources and support than whatever faculty member had time to be their adviser. As far as academic committees go, it’s unrealistic to expect that we have a committee for each of the 4,000+ impacted populations we are reporting on, but it’s not unreasonable to have a more intersectional set of third-party experts included in the process of addressing impacted success.

Let’s quickly illustrate a far-superior alternative to the current system. Imagine each of the current committees had a member tasked with working on each of the eight demographic dimensions. So now there is at least one person on the Gender Equity committee who is monitoring the success of each ethnic group. For example, not just women but black women and latin women and pacific islander women, etc. Another committee member could be charged with looking at Foster Youth status; how are women who were foster youth doing as compared to those who were not foster youth? Discussions between these two members would be far more attentive to intersectional challenges (while also suddenly addressing all ten of the most impacted populations) than the current model where both members are talking about women in the abstract, and potentially ignoring the fact that there is no woman who is without race or foster youth status.

Opportunities For Further Work

Perhaps the most important piece of future work on this topic could be incorporating more demographics data about student populations. In many cases, Sierra College is not even tracking critical demographics data about student populations. For example, a student is only counted as disabled if they ask for an accommodation which is subsequently approved by the college. A student who shows up for class in a wheelchair while struggling with severe depression but does not ask for any accommodation (or whose request for accommodation is denied) is not counted as disabled according to Sierra College. This needs to change.

Additionally, Sierra College is not currently reporting all the required success metrics as listed by the California Community Colleges. Specifically, there are enormous swathes of data missing from the public website. In some cases, metrics are omitted entirely. For example, there is no mention of the required metrics “number of students who complete nine or more career education units,” or, “number of student who have attained the regional living wage.” In other cases, entire years of information are simply left blank where the data is reported at all.

Additionally, many longitudinal metrics are falsely reported as static by the school. For example, we have no way of determining the success of students who come out as LGBT at Sierra College because the school does not update this critical longitudinal demographic data. Instead the school falsely reports these longitudinal values as static when students enroll, as though these metrics do not change. So a student who comes out as gay or transgender at Sierra will be counted as straight or cisgender for their entire career at Sierra. Since these are already deeply impacted populations, their suffering and marginalization is going completely unreported by the school. It’s not clear to what degree this false and inaccurate reporting may be deliberate. The problem has been raised by interested parties and stakeholders on multiple occasions in recent years, to no change from the school.

Filling in these data gaps and correcting deeply unjust metrics would make important work for future researchers and activists at Sierra College.

Lastly, the work of updating and publishing this data should be conducted by the school’s research department. Students should not have to take unpaid internships and work for months in order to answer basic research questions which the school should already be reporting through its well-funded research department which ostensibly exists for this purpose. Because of the claims made my the research department that this sort of work is unfeasible, and the fact that a student was able to conduct it without any internal access based solely on public data, I have also included all the tools I built to conduct this research so that there is no further excuse not to do this, and to allow other activists and students to conduct further research of their own.

Conclusion: My Suggestions

  1. We need more resource centers and engagement centers which are staffed more consistently and are available for students who need them.
    1. We need more resources and better training so that staff can actually explain them to people who need them.
  2. We need independent third party experts to be invited and welcomed and included to participate in engagement centers and student organizations as well as academic senate committees and especially equity committees.
  3. We need better outreach and research to honestly explore where the problems are and what the best solutions are.
  4. We need more intersectional cohorts. (ie. black foster youth, queer people of color, etc.) Many of the groups identified in this meta-analysis are more than large enough to form a cohort. This would allow students in these populations to build community ties with people like themselves. This is critical for success.
  5. We need to be funding equity committees and giving them actual power and permanent staff. These problems are not going to be solved by inviting untrained staff to volunteer on a toothless and unfunded committee.
    1. We need more equity committees. Having at least one for each of the demographics we are required to report would be a good start. These committees should have members looking outside their core demographic at other demographic intersections which constitute the many populations under the umbrella of their core demographic.
    2. Equity committees need training on intersectionality and on actually impacting success rather than simply engagement. These committees need to be working together and meeting more than biweekly, and doing more than planning fundraisers.

Data: Are Shiftpods Actually Cooler?

My Findings;

  • Shiftpods are not significantly better than cheap knockoffs.
  • No tent performs well in direct sun.
  • No matter what kind of tent you have, putting it under a shade structure which is at least a foot away from the tent is the most important thing for keeping it cool inside during the day.
  • Evaporative “swamp” coolers are not as helpful as simple ventilation.
  • Thermal mass batteries are a simple and very effective tool for normalizing temperature.

The Experiment

Lots of people spend time in tents. As a Burner, I am one of them. There is a big problem which anyone who has ever slept in a tent can relate to. That problem is Thermodynamics.

Thermodynamics is a law of physics which basically says that the heat from the sun will eventually get into your tent. What this means for campers is that even on a day where the outside temperature only gets to 80, it’s going to be 100-120 degrees inside your tent. There are basically two ways to fight back against thermodynamics in this case; put barriers in the way which will slow but not stop the heat getting in, or put something in place to absorb or remove the heat.

For a Chemistry honors project, I am studying the effect of various techniques and materials on prolonging that magical sleeping time in the morning before the temperature inside the tent gets too high to keep sleeping.

I will also be comparing fancy expensive tents to the cheapest alternatives in order to see if fancy expensive tents are actually any better. In this case, I have chosen the very popular Shiftpod v2.0 to compare with a cheap tent of similar design from amazon. I have taken to referring to the blue fishing tent as “Blueshift” for simplicity’s sake.

One thing to keep in mind is that this is not the playa. We do have partly cloudy days here in Sacramento, so there may often be sudden spikes and dips in temperature. These cloud effects will reach both tents, so it will be interesting to see how they each fare with them.

Disclaimers

This experiment was done in Sacramento, not at Burning Man. Results will vary but systemic error should be low, since both situations include similar variables such as humidity, sunlight, etc and impact from the different mitigation strategies. I’m planning to do a future version of this experiment at the playa during fourth of juplaya which includes more tent types such as a hexayurt and a pup tent.

The tents are not exactly the same size. There may be some differences in amount of heat reaching each tent. That said, Shiftpods are 1200% the price of the cheaper tent. If it’s worth that, this trivial difference should not matter.

Some of the links on this page are affiliate links. I”m a starving student trying to pay for this expensive experiment. It doesn’t add to your cost if you choose to buy things with these links. I just get a small cut to help cover my costs. If you don’t want to use the affiliate links, you can open them in incognito or something like that in order to remove the tracking code.

Sensor Equipment

I am using a custom built multi-zone temperature logger to track the data. The full details of that can be found here.

Phase 1: The Naked Sun

For phase one of the trial, both tents will be in full sun. This photo shows them at 2pm the day before the data shown. They had a full 24 hours to adjust to the environmental conditions before experimental data collection began.
Like any tent, both the Shiftpod and the Blueshift will eventually get much hotter than the outside temperature when in full sun. The way to win at this stage is by staying cooler for longer. This means the occupant will be able to sleep longer and have a better experience with their tent.
In fact we see that the internal temperature of the Shiftpod hits 100 degrees more than an hour before the Blueshift does. At 9:30 in the morning, that crucial time for sleep, we see the Shiftpod more than 15 degrees warmer than the Blueshift. By 10:45, it has passed 100 degrees. The Blueshift doesn’t hit 100 until after 11:45.
Phase 1 Graphics
At this point, I suspected a possible reason for these surprising results. The Blueshift has a single layer of thin waterproof fabric, where the Shiftpod has a thick layer of air-tight insulating fabric. I suspected that this thick insulation could be trapping the heat where the Blueshift will let that heat escape.
This could explain the difference in temperatures, and imply that Shiftpod themselves intend users to take additional measures in order to reap potential benefits of the materials they use. Air conditioning for example would make that insulation helpful rather than harmful. This would also explain why Shiftpod has recently started giving away free “blast shields” or radiant barriers to go on top of the Shiftpods.
I decided to test this theory, so I added another phase to the experiment to compare the difference in temperature inside the tents from floor to ceiling. Here is what I found;
Phase 1 Adendum
As you can see, the temperature between the floor and ceiling in the Blueshift stays almost the same throughout the day. In contrast, we see the heat being trapped inside the Shiftpod. The temperature difference between the floor and ceiling of the Shiftpod varies by as much as ten degrees in the hours leading up to one-hundred degrees. This is very interesting information because if we are able to deal with that trapped heat, we could potentially see benefits from the Shiftpod’s insulation. This will be addressed in an upcoming phase (Phase 4) of the experiment.
 
In Naked Sun, the Blueshift is the clear winner for morning temperatures. This is the time and setting where the Shiftpod is supposed to shine. Instead we see it falling  behind its much cheaper cousin.

Phase 2: Loose Aluminet

Aluminet is a very interesting material which has found wide adoption among burners, just like Shiftpods. It’s both a radiant barrier and wind-permeable. This solves several interesting problems.

The idea of a radiant barrier is that heat and light will be reflected rather than absorbed, as would be the case with a tarp. The problem with past radiant barriers is that they were solid like a tarp, and with high winds, they become a sail that tries very hard to blow away. Many radiant barrier materials such as mylar are very delicate and will tear themselves to shreds with even the slightest wind.

Aluminet is made of a similar material to mylar, but it is reinforced and braided, so it forms a very strong barrier in the wind, which wind can actually pass through. It comes a wide range of opacities which allow different amounts of light through. I go with 80% Aluminet.

As I mentioned before, Shiftpod has recently started giving out free radiant barriers called “blast shields” which are basically just a non-breathable mylar -like material. The blast shield is basically a cheap and inferior alternative to Aluminet, and the fact that they are now giving them away for free seems to indicate that the makers of Shiftpod are aware of a potential design flaw in this very expensive tent.

Thermodynamics sometimes has a problem with both mylar and Aluminet, especially when they get dusty. Sure radiant barriers will reflect much of the light and heat, but dust will not. If the dust is touching the radiant barrier, then it doesn’t stop the heat from the dust conducting through the barrier and into the tent. Aluminet and mylar are not conductive barriers, only radiant barriers. This means that a dusty piece of Aluminet or mylar will conduct some heat through to anything they are physically touching. For this reason, if the radiant barrier is touching the tent, and it’s dusty, then it will not block as much heat as it would if it were several feet away from the tent, suspended from another structure. This will be expanded on in a future phase of the experiment.

Phase 2 Photo

This photo was taken in the afternoon. As you can see, the Aluminet is hanging from the two tents and tied to the fences. For the purposes of this experiment, the Aluminet is covering the east sides of the tents. This means it will block morning sun, and to a lesser degree the noon sun. The south and west sides of the tents are not covered. (This entire experiment focuses on the morning time.)

Phase 2 Graphics

The addition of Aluminet pulls the curves closer together, and we see the Shiftpod performing much better after an initial heating-up period. But the thickly insulated Shiftpod is still much warmer than the thin and breathable Blueshift. The Shiftpod has two humps when the sun is hitting it straight on, first thing in the morning and then again at mid-day. Between these times, the sun is hitting its edges at an angle. This seems to impact the heat absorption as shown above.

Once we pass noon, the Shiftpod’s thickly insulated roof absorbs the conducted heat from the loose Aluminet, and traps that heat inside the tent. In contrast, the Blueshift’s breathable and un-insulated roof lets the heat escape. Right at noon, we see the Shiftpod suddenly jumps to nearly twenty degrees warmer than the Blueshift. This could be the problem that the makers of Shiftpod were trying to solve by giving everyone free blast shields to block the mid-day sun from cooking their thickly insulated roofs. By comparing the first and second phase data, we see that adding a radiant barrier op top of the Shiftpod does mitigate this problem to a certain extent, but not completely. Eventually the heat gets through since the barrier is touching the roof.

Phase 3: Passive Thermal Mass Batteries

Water as a molecule has a lot of interesting properties. One of these properties is that it’s really good at storing heat energy. In fact, my understanding is that no known material has a higher density of heat storage than water.

This gave me the idea of placing lots of water inside the tent to store and release heat. This would theoretically have the effect of flattening out the temperature curves so that they are closer to the average value throughout the day. I decided to use four five-gallon buckets of water as my passive thermal mass battery. These buckets would be easy to transport and store, and even use as storage for other gear. You could also fill one or two of those large 27 gallon totes and use that as your battery, but that seems like overkill and it would be a lot of water.

As you can see in the photo below, I also put one of my temperature probes into one of the buckets to monitor the temperature of the water so we will see that in the graph.

In order to accurately compare the performance of the two tents, I decided to split this phase into three parts and compare each tent’s performance against the other, then both together. Therefore Phase 3A will test the performance of the Shiftpod with a 20 gallon passive thermal mass battery. During this phase, the Blueshift will just be normal and without any thermal mass battery. Then during Phase 3B, they will be switched. The thermal mass battery will be placed in the Blueshift, and the Shiftpod will just be normal. Then we can compare the performance of each with the addition of the batteries. Lastly, during Phase 3C, both tents will contain the same thermal mass batteries.

In order to get the most accurate data, I let the tents and their thermal mass batteries sit for a full 24 hours before collecting the data as shown. This way we can see the data once the batteries have been well adjusted to the conditions we are observing.

Phase 3A: Shiftpod Only

Phase 3A Photo

Phase 3A Graphics

I was a little surprised by this data. In the first graph above, we see that the thermal mass battery starts absorbing a great deal of heat long before we see a significant change in air temperature inside the Shiftpod. To really see the impact, compare the difference chart to previous phases of the experiment. Only in the pre-dawn hours does the Blueshift take the lead. I am amazed. With the simple addition of a passive thermal mass battery, the Shiftpod now performs better than the Blueshift all throughout the day.

Phase 3B: Blueshift Only

Now let’s look at the next part of the experiment, where the passive thermal mass battery is placed only in the Blueshift and the Shiftpod is just normal without any help from the batteries.

Phase 3B Graphics

We see that the difference curve swings the other way. This is not surprising since the Blueshift is not insulated. I think this test should be compared to the first test in the experiment since that’s the same conditions the Shiftpod is under during this test. In the first test, both tents eventually reached essentially the same temperature, and the Shiftpod was slightly warmer for most of the day. In this test,we see that both tents follow very similar curves, but that for most of the day, the Blueshift is a good 10-12 degrees cooler than the more expensive Shiftpod.

It’s interesting to note that we don’t see precisely the same effect from the previous test (3A) where the thermal mass battery absorbs the heat that would otherwise constitute the curve of it’s surrounding tent. Instead we see a gradual warming throughout the day, and a reduced warming in the surrounding tent.

Phase 3C: Both Together

Phase 3C Graphics

During this phase of the experiment, we saw that thermal mass batteries significantly improved the effectiveness of the Shiftpod, while not significantly improving the effectiveness of the Blueshift. Four buckets of water has a footprint of just a few square feet. You would barely notice these in the tent, and they seem to provide an enormous benefit. Probably you would see similar effects from filling an empty tote with water. This would be an interesting topic for a future experiment.

Phase 4: Active Thermal Mass Battery

In order to increase the efficacy of the thermal mass battery, it seems logical to test adding a fan, pump, and heat sink. This would mean you would need to supply constant power to it, but if it provides enough of a benefit, then this may be worthwhile. I removed three of the buckets for this phase of the experiment, using only five gallons of water in a single bucket. Expanding on this may prove worthwhile but at this point, connecting buckets would be complex. If we wanted to do a larger active battery, then probably a larger container would be more appropriate. This would also allow you to add ice to your container, etc in order to cool the tent even more.

Hanging this from the ceiling also has the added benefit of potentially mitigating that pesky heat-trapping effect in the Shiftpod.

Phase 4a Photo

Pictured above are a 240mm heat sink with two c. I hung the heat sink and fans from the roof of the Shiftpod and put the pump and hose into the bucket. Here is what it looked like in the early evening. (You can see the sun reaching behind the Aluminet to hit the Shiftpod from the afternoon side. This time of day is not part of the experiment.)

Here is the data on the active thermal mass battery in the Shiftpod only;

Phase 4A Graphics

As you can see, the smaller thermal mass battery just isn’t able to keep up. Being active seems to make little difference in terms of absorbing the heat in the tent. This is an unexpected result; very interesting data.

Because we see such a limited effect even with the insulated Shiftpod, and because the larger thermal mass battery had essentially no effect on the Blueshift, it does not seem worthwhile to test an active thermal mass battery in the Blueshift.

Phase 5: Ventilation

As you can see in previous experiments, the temperature inside the tent can be as much as twenty degrees hotter than the outside temperature for much of the day. This suggests a simple solution; ventilation. This can be a problem in dusty environments at Burning Man. In other environments, this may be a better option.

Ventilation & Dust mitigation

I did some experimentation this year at the burn based on Nikola Tesla’s patents for solid state check valves as well as the closely related concept of vortex filtration. The basic principle is that when you have a fluid such as the air, and there is something in suspension, such as the dust, you can encourage the dust to fall out of suspension by forcing the fluid to suddenly change directions.

My idea was to take a section of about eight or ten feet of aluminum duct and hang it so that at least a few feet were hanging loosely and free to flap around in the wind. The other end of the duct would go into the tent, where a powerful fan would suck air into the tent. I set up some other ducts (pic) to maintain constant slightly-negative pressure and pull the hot air out from the top of the tent. This meant that no dust would blow in through these other ducts. The theory was that because the outside piece of duct was free to flap around, the open end would never be facing into the wind. This means that when air is sucked into the duct and drawn into the tent, that air would have to change direction 180 degrees in order to get into the duct. This means the dust would fall out of that air. As far as I can tell, this worked perfectly. I would definitely recommend trying this at the burn.

Phase 5a: Ventilation (Shiftpod Only)

The basic idea is pretty simple. We will have a fan drawing the air out of the top of the tent at 200CFM 24/7. Then an open vent will allow fresh air to enter the tent. This means all the air in the tent will be replaced every five minutes. This is generally considered to be the ideal ratio for ventilation, especially with evaporative cooling as we will see in the next experiment.

Phase 5A Photo

As you can see above in the photo, I have used the same hose from the previous experiment as well as the same 200CFM fan. The direction of the airflow is out of the tent. You can see the duct drawing the hot air from the roof area of the Shiftpod and then pushing it outside the tent. A window is open to allow fresh air to enter the tent, replacing the hot air being pushed out. This photo is for simplicity’s sake, but having the open window be close to the exhaust duct doesn’t make sense. In the actual experiment, only the window opposite the exhaust duct was left open.

Phase 5A Graphics

As you can see, just like the passive thermal mass batteries in Phase 3A, ventilation swings the difference curve in the direction of the Shiftpod. Also just like Phase 3A, this puts the Shiftpod ahead by about ten degrees. Recall that with no assistance, the Shiftpod is a few degrees warmer than the Blueshift. Knowing that these two methods each impart a ten degree advantage plus the unaided difference, we will likely see large advantages when we pair these methods together in a later trial.

Phase 5B (Blueshift Only)

This is the same experiment with a 200CFM fan drawing the air from the center of the ceiling of the Blueshift and blowing it outside while fresh air comes in from another vent.

Phase 5b Photo

Here you can see the same fan and ducts going out through a window which is covered with cardboard for this experiment. Let’s see the data…

Phase 5B Graphics

As you can see, adding ventilation did make the Blueshift cooler than the Shiftpod, but only by a few degrees. I was surprised to see how little impact was measured in phase 5 of the experiment.

Phase 6: Hung Aluminet

From the beginning, I suspected this would be the most impactful phase of the experiment. The idea here is that the Aluminet is suspended a few feet away from the tents in order to eliminate any potential heat conduction between the radiant barrier and the tents. This seems like the most logical step to take in order to see a really dramatic change. This is where the Aluminet will really shine.

Flat top shade structures are expensive! But they don’t need to be. Check out this post about how I cut the cost in half and eliminated most of the hardware involved while still getting all the benefits of a normal flat top shade structure.

Phase 6 Graphics

As you can see, this material has a big impact versus the previous phase. The Shiftpod never hits one-hundred (as an average of its floor/ceiling temperature), though the temperature at the ceiling briefly does. The radiant barrier on the Shiftpod seems to really shine in this phase, with the highest temperature difference between Blueshift and Shiftpod we’ve seen so far.

Phase 7: Evaporative Cooling

Evaporative cooling (AKA swamp coolers) add humidity to the air in order to reduce its temperature. This works best in dry climates. The maximum temperature difference is a function of the dew point under the given conditions. This means the best case is a 40 degree temperature difference if there is 0% humidity, and less of an effect with higher humidity, with no effect at 100% humidity.

For this phase, I used the evaporative coolers I designed in this post. In both cases, the coolers burned about three gallons of water during the period shown.

Phase 7A: Shiftpod Only

Phase 7A Graphics

Based on comparison of the data from Phase 6 to Phase 7A, there does not appear to be any significant temperature difference in the Shiftpod with the addition of the evaporative cooler. On the day the data was collected, humidity was 68%. This is probably why there was no significant temperature difference. Despite that, the evaporative cooler did burn through three gallons of water during this period, so it was working and cooling the Shiftpod despite the unimpressive results.

These values would likely be different if there was a significantly lower humidity. This was a problem at the burn this year, where the humidity was uncharacteristically high, and evaporative coolers were not working very well.

Phase 7B: Blueshift Only

Phase 7B Graphics

As I suspected, evaporative cooling was basically useless in the Blueshift as well. This matches my experience from on-playa this year. By the end of build week, I threw my swamp coolers in my camp’s dumpster.

Phase 8: All Or Nothing

In this final phase of the experiment, each tent will take a turn with all or nothing. Based on comparing the results of Phases 5 and 7, there is no reason to include useless evaporative cooling in this phase. Instead, the more compelling phase 5 results will be included and simple ventilation will take the place of evaporative cooling. The more I think about it, the more sense this makes. If it’s twenty or more degrees cooler outside than inside, then simple ventilation is guaranteed to accomplish that or something close to it. This also parallels my playa experience, where ventilation is often the best and easiest option.

Phase 8A: Shiftpod Only

Phase 8A

For this phase, the Shiftpod will be under hung Aluminet with the 20 gallon passive thermal mass batteries, and ventilation. The Blueshift will have none of these things, and be sitting out under direct sun.

Phase 8A Graphics

Here we see the effectiveness of the ventilation system sucking the hot air out from the top of the inside of the Shiftpod. For much of the day, the ceiling is actually cooler than the floor.

Phase 8B: Blueshift Only

For this phase, the Blueshift will be under hung Aluminet with the 20 gallon passive thermal mass batteries, and ventilation. The Shiftpod will have none of these things, and be sitting out under direct sun.

Phase 8B Graphics

As with several previous phases, there seems to be a critical tipping point when the Shiftpod suddenly becomes much warmer. My going theory is that it is related to the angle of incidence of the sun. I found the results of this last phase very surprising. I would have expected the difference curve to be much flatter and to stay below zero all day.

Final Thoughts

In the naked sun or with loosely draped Aluminet, it’s actually hotter than the cheaper Blueshift. Both tents perform a lot better under a shade structure.

Simple ventilation is more effective and comfortable and less complicated than evaporative cooling.

The Shiftpod is over ten times the cost of the Blueshift. Does it perform ten times better? No. In general, the tents perform pretty similarly. The Shiftpod excels in a few situations with some help. If I was choosing a new tent today, I would pay more attention to getting a good shade structure than to getting any certain kind of tent.

DIY High-Precision Multi-Zone Temperature Logger

I was in need of high-precision multi-zone temperature logging gear for a chemistry honors project, and I found nothing online that would work well for my purpose, so I decided to put my CSCI46 skills to use and build my own.

The Hardware

I started with a NodeMCU which is basically a fancier but somehow also cheaper Arduino alternative which has wifi built-in. It works just like Arduino and you use the same Arduino IDE to program it.

I decided to use DS18b20 temperature probes which come in a waterproof five-pack with long cords. These also need a 4.7k ohm pull-up resistor which I will explain in a moment.

So one really cool thing about these temperature probes is that they work with the OneWire library. This means you could use hundreds of these probes and they would share only a single I/O pin on the Arduino. My implementation uses five, but you could easily use many more if you wanted to.

Expose about 1 cm of wire

Each temperature probe has three wires coming out; red, yellow, and black. Start by trimming the colored portions a bit so that there is around a centimeter of wire exposed. Then zip tie the probes together as shown, so that they won’t pull apart when in use. Next, twist each color together as shown. I used like-colored jumper wires to create clean and simple pins to plug into the Arduino, but however you want to accomplish that is fine;

Twist similar colored wires together

The next step is to incorporate the pull-up resistor. This needs to go between the red and yellow wires. I added extra zip ties just to be safe.

Pull-up bridges yellow and red wires.

Lastly, simply tape all of that up so it doesn’t come apart. My final product looked like this;

Final Product

The Code

You will need to download and install the Arduino IDE, the Temperature Sensor Library, the OneWire Library, and the CH341SER USB Driver for the NodeMCU Arduino. Then go into the IDE and add the following URL in the preferences section as an “Additional Boards Manager URL…”

http://arduino.esp8266.com/stable/package_esp8266com_index.json

Next, take a look at my final code, here;

https://github.com/cjtrowbridge/Sensor-Server

Simply add your wifi ssid and password, and the code should work as-is. It will output the IP address it has gotten from the router to the Arduino console. A best practice is to simply set this as a static IP on your router. Then, it will not change, and you will always be able to find the device.

The Data

Navigating to the URL of the device’s IP will present a comma-delimited, quotation encapsulated set of values for each sensor. There is no way to tell which sensor is which based on the hardware. The sensors have a built-in serial number, and the OneWire library knows which sensor is which. The list served at the URL will always show the values in the same order on that basis.

So we simply need to determine which sensor corresponds to which value. You will need to observe the sensor data and then place a warm hand on each sensor, noting which value changes temperature. I used tape and a sharpie to label each cord (1,2,3,4,5) so I would know which sensor corresponded to each value in the list the device serves at its URL.

Lastly, I wrote a cron script which runs on my Synology NAS and fetches the values from the device every minute, and writes them to a CSV file. This means that I can later look back and simply import the values into an excel spreadsheet.

Cheap Flat-Top Shade Structure (Alternate Design)

Shade structures are perhaps the most important and powerful way of improving temperature conditions in a tent on playa or elsewhere. The value of a shade structure to someone who is trying to sleep can not be overstated.

The Traditional Way

A lot of people buy and build complex and expensive shade structures. Typically these are made using five-way joints with feet and ten foot sections of one inch emt conduit (or here). These are typically covered with tarps and ratcheted down.

Traditional flat top shade structure

Parts List

10′ by 20′ Total Cost: $382.98

Another Approach

At the burn this year, my friend Giorgio suggested putting up a flat top using rebar instead of feet and aluminet instead of tarp. This would mean far fewer parts, no need to ratchet strap it down since it won’t catch wind, and much easier to transport. I used paracord to tie down the aluminet which worked perfectly.

I simply pounded rebar into the ground, leaving about a foot above ground. I tied paracord to the eye bolt and dropped it through the aluminet’s grommet and then into the end of the emt conduit. This is fine since the cord is not bearing any load. It’s basically just there to keep the eye bolt from falling out if the aluminet flaps a bit. The wind blows straight through the aluminet so the force on the paracord is minimal. The other end of the conduit goes onto the rebar. The paracord is tied off on some lag bolts (This is overkill. There is probably a more elegant and less complex solution here). Below is a picture of this set up in my back yard as part of my chem honors project.

Parts List

10′ by 20′ Total Cost: $158.29

Less Than Half The Price

Wow so this design is less than half the price of the traditional model and comes with far less to haul in to camp. I will definitely use this design from now on! Here’s what it looks like in my back yard…

Shiftpod and Blueshift Under Aluminet

Psychedelic Glasses From Decomp ($10)

Burning Man Decompression was a blast this year. I wore a pair of psychedelic glasses which made everything look like this. A lot of people asked me for the link to them, so here it is!

They just look like normal glasses from the outside…

Psychedelic Glasses

Another person had this pair which were also popular so  decided to attach the link to them as well.

Another pair of psychedelic glasses

My 2019 Takeaways

The blue knockoff shift pod was an amazing success. It was so easy to set up and take down. It fits perfectly under a flat top shade structure and it stays cool until after noon. Along with my favorite ear plugs, the blue shift pod allowed me to get more sleep than ever before at a burn. I can’t recommend these highly enough.

I brought a notebook this year so I could write down things I wanted to change or bring next time. Here are some of the things I wrote down;

Bring More Fans: I brought a pair of high speed fans for circulation. When you’re under a flat top shade structure, often times the air outside stays pretty cool. Pumping that in can make a world of difference. And it augments the effectiveness of the swamp cooler. I set up two 200CFM fans with ducts to draw air in from outside. These were on a simple thermostat. They automatically kick in when the temperature starts to rise. I estimated that adding these simple circulating fans gave me about two extra hours of cool temperatures for sleeping in. In the future, I think I will double this set up and bring four of these 200CFM fans while ditching the swamp cooler altogether. The swamp cooler is really large and complicated and needs a lot of water and isn’t super effective. Simple circulation feels like a lot of value for not a lot of money and space, in my experience.

Bring A Parasol: I felt like the only person without a parasol this year. I borrowed one from a friend in order to go to the DMV with another camp one day. Having that tiny bit of shade to bring with you on playa makes a world of difference.

Flat Top Shade Structure: This year, my friends and I brought several large pieces of aluminet. Initially, this was just laid over the blue shiftpods. This did not help at all with the heat. In fact, it created lots of noise flapping around, while not helping with the heat. Our tents were right next to a 20′ by 96′ flat top shade structure which formed the core of our camp’s residential area. I experimented with moving the aluminet to attach to the shade structure and then to poles on the other side of our tents. This means that the aluminet was hanging taught a few feet above our tents, rather than touching our tents. This made a HUGE difference with the heat. That said, mounting the far end of the aluminet to poles stuck into the ground was not ideal because they moved around even though they were guyed down. I  think a more ideal system would be to set up another small flat top twenty feet from our main flat top, then stretch the aluminet between them, and set up our tents underneath. This would mean a lot of extra shade for super cheap. #lowhangingfruit

Failover Switch: My camp provides power to every tent. This is great, but if you’re relying on it for something like ventilation and then it goes down, you will get hot and wake up. I also brought a large battery, but it doesn’t charge super fast. Using it as my main power source would not work. It would discharge overnight and then not fully recharge in time to use it again. The solution is a failover switch. I have actually already written a proposal for using these for our kitchen refrigerators. The idea is simple; a battery charger and a failover switch both plug into the main grid. The battery charges from the grid (or from solar or wind). Then, the failover switch provides power from the grid to all your devices while the battery is charging. If the grid fails, then the failover switch immediately switches to drawing power from the battery. The devices plugged into it don’t even notice. When the grid comes back on, the switch immediately moves back to drawing power from the grid and the battery begins to recharge. This is especially important for ventilation and air conditioning. One morning this burn, the power failed and the tent quickly heated up to an uncomfortable temperature, waking me up. Never again!

Renewable Energy Sources: With the layout described above, the battery can easily charge off of solar or wind rather than the grid. Then, the switch could be hooked up in reverse so that we first use the stored solar power, and then switch over to using the grid. This would offset diesel fuel costs, and reduce our carbon emissions!

Pens, Sharpies, Calling Cards, Ziplocs: Ziplocs are perfect  for storing the many small items which accumulate during the burn. There are many cute examples of calling cards which my fellow campers were giving out this year. Pens and sharpies are extremely handy, especially during strike.

Snacks: My favorite two snacks this burn were frosted strawberry pop tarts and peanut butter filled pretzels. Specifically we had the giant kirkland buckets. I will definitely bring some of these next year.

 

Advanced Playa Air Conditioner

I did another post about building a simple playa air conditioner. I made several major innovations over existing designs, but also repeated a major failure of existing designs. In this post, I will describe my process of addressing this failure and improving further on the prevailing designs for DIY playa air conditioners.

Fundamentals

All DIY playa swamp/bucket air conditioners work on the same principle called evaporative cooling. The basic concept is simple; you force liquid water to evaporate into dry air. This turns hot, dry air into cool, humid air.

All the designs that I’ve seen use evaporative pads. Water is pumped over the pads, saturating them. Then a fan pulls air through the pad, evaporating the water in the pad. The cold air coming through the fan is ducted into a tent or yurt.

The Problem

Mots of these designs, including the apparent original, share a common flaw; they expose only a few inches of the pad to the outside air. These designs typically feature a bucket with a few small holes at the top to let air in through the pad.

The first problem is that the pad is designed to work across its entire surface area, and exposing just a few square inches of it to the outside air means the pad is not really working as well as it could.

The second problem; the holes are at the top because the bucket is full of water. So even if the pad was fully exposed to the outside air, most of it would be under water.

My First Design

My earlier post featured chicken wire which holds the pad away from the edge of the bucket. This means that the air from outside is able to reach the entire surface area of the pad.

The problem with this design is that the bucket is still full of water. So even though the pad is fully exposed to the outside air, most of the pad is under water.

Version Two!

Here are the two buckets with their lids side by side. The top bucket is the evaporative chamber. The bottom bucket is the water reservoir. Cutting through these is easy with a simple box cutter.

 

As you can see, the first bucket has drain holes which lead down into the reservoir. The evaporative pad sits in a column under the duct, drawing in hot, dry air from all directions.

 

Here is the evaporative pad, surrounded with chicken wire and with two drip lines installed. These are just sections of tubing with their ends crimped with zip ties, and holes punched every inch or so with a hot soldering iron.

 

When the evaporative column is inserted into the top bucket, the lines run down into the reservoir bucket like so.

 

I replaced my previous pump design with this one which is more clean and probably more reliable. These pumps have barrel jacks which run through a splitter and then to a power supply. Also, these are beefier pumps because the USB-powered ones could not lift water up high enough for this design.

 

I also found a beefier 200 CFM fan which can run into the same power supply. I added one duct adapter to each side, and attached an eight foot section of 4-inch dryer duct. This will run into the tent. The other good thing about using barrel jack power is that I have several batteries which will be able to run this newly redesigned system.

 

The same dryer duct quick disconnect runs down into the evaporative column and connects to the fan and duct.

 

This design improves on all the problems identified with the previous design. Looking forward to a cool playa experience!

 

Parts List

Anti-Darkwadding with Sound-Reactive Lights

I’ve been talking about doing this for a while and I finally pulled the trigger. I will probably safety pin these to bags and costumes throughout the burn. I think these would also be cool inside a clear backpack. Check out the video!

Parts List

3x USB-C Light-Up Charger Cord

3x USB Sound-Reactive Disco Light

3x USB-C/A Male/Female Dongle

3x USB-A Female/Female Dongle

1x USB Battery Pack

This is the battery pack that I used for the video. I tried measuring the power consumption of these lights but it was too low to register on my measuring device. Probably this battery is plenty to power these lights for the whole burn. And it’s easy to recharge with a solar panel like the one I’m bringing.

The dongles allow the USB-A Male of the disco light to connect to the USB-C Male of the charging cord like so;

 

Queerborhood Wifi

Phase-One

This is the final draft proposal for the pilot project to tie into the new borg network. If successful, the project could eventually deliver WiFi to the Queerborhood (7:30 sector). There have been several changes of priority since the first draft of this proposal. This final version features a mast with a microwave dish to tie into the borg network, a wifi router, and three range extenders which distribute the connection throughout our camp. This way, later we can add more extenders as well as load balancers in a future phase. This is intended as a first step in a larger long-term project.

For a brief description of my relationship to this topic; I am a licensed FCC Radio Tech (KK6VJX) with two decades of experience in supporting computer networks.

Normally, we would need to install enormous masts to connect to the borg network, but luckily we have three semi-trailers which stay in the camp throughout the burn. This means we can place smaller masts on top of the trailers in order to tie into the network. Specifically, I am recommending a “non-penetrating” mast designed for flat roofs. Instead of weighing it down with bricks, we will use two pairs of three ratchet straps to tie it around the trailer.

Per the borg’s recommendation, the main connection is a Ubiquiti NanoBeam microwave dish. This is wired into the junction box on the mast where a cheap wifi router takes the connection from the microwave dish and distributes it via wifi to the camp.

Throughout the camp, cheap wifi repeaters extend this signal to make sure our network reaches the entire camp. These will be included in the junction boxes of our power grid’s new distribution substations. During the proposed second phase of this project, additional masts will be placed in various locations and tie together via a load balancer which then feeds into our wifi. This means that if there is any problem with any of the masts, the other ones will continue to work. At that point, any neighboring camps that want to tie into Queerborhood.net will simply need to use those cheap wifi extenders to tie into our system since we are doing all the heavy lifting of connecting to the borg network.

Topography

 

Parts List For First-Phase

Learn To DJ: (4) Getting Gigs

2019-07-26

6:30-7:30

Are you DJ-Curious? Comfort & Joy is hosting free DJ workshops for anyone interested in learning this magical art. The workshops will be led by seasoned disc jockeys from very different parts of the DJ industry. You will learn how to set up sound and lighting gear, how to mix music, and how to pick tracks.

Food and drink will be provided. Anyone who attends all four workshops will receive a very special gear gift from C&J.

At this fourth workshop, we will learn about the history of the vocation of DJing, and how to find work today. We will have several local big names on hand to answer questions about the work itself and finding gigs.

Workshop Outline

  • Meet and Greet
    • “What kind of gigs are you interested in?”
  • History of DJing
    • Challenges and advantages
      • Then: Heavy, expensive crates of music
      • Now: Cheap controllers with MP3
    • Client expectations
      • Are you video mixing?
      • Are you bringing lights?
      • Are you taking requests?
    • Selling gigs
      • Schools
        • Find the decision maker
        • Keep bugging them
        • Sign exclusive seasonal contracts; offer free work as an audition with the option to cancel the contract if they don’t like it.
        • Play top 40/ contemporary
          • Have edited music crates
        • Typical pay is about $100/hr.
      • Weddings
        • Mostly word of mouth
        • Start with friends/ family
        • Win over wedding officiants
          • Make a special effort to network with them at weddings
        • Talk to wedding venues about being their recommended vendor
        • Typical pay can be $600-$1200.
      • Dive bars
        • Easy sell, low hanging fruit
        • Check local newspapers for bars with empty nights. Ask for a tryout.
        • Typical pay is a percentage of sales or a flat rate: $100-$200 plus a free bar tab for you and your guests (we’ll come back to that).
      • Clubs
        • Need to know someone or have a reputation
        • Pay ranges wildly from minimum wage hourly to high flat rates depending on the venue and the night.
      • Look for weekly/ regular gigs!
        • One-off parties are fun, but weekly gigs offer stability and reliable income.
        • Working three nights a week at dive bars pays about $20k/year and sets you up for bigger things.
        • Weekly dive bar gigs are a great reputation piece for selling larger gigs.
        • Exclusive contracts with schools are a great reputation piece for selling more school gigs.
    • Guerrilla marketing
      • Flyers only help if they come at the end of a conversation.
      • Talk to people. Invite them.
      • Go visit similar events.
        • Set up collaborations
        • Invite guest DJs to spin a few tracks during your sets while you take a break. Buy them a drink on your free tab in exchange for them telling people to come to your party. Check in and tag them to reach their following online.
      • Have a larger cause/ purpose
        • Nonprofits will bring enormous numbers of people to your party if they get a cut of the sales.
        • People feel better about partying if it’s for a good cause.
    • Q&A Panel
    • Mixer