The Hidden Evils of Data Centers

“I want to see criminals behind bars!  A jail is being constructed down the road.  Don’t you put that jail near me!  I don’t want a jail in my community!  The traffic, the influx of visitors, the pollution, the sewage … all of it will poison my community.  It will devalue my home!  I want to see criminals behind bars, just put them somewhere else.”

This type of rhetoric goes on everywhere.  With every part of the public service world.  Jails, highways, shopping centers, low-income housing, hospitals, prisons … and now data centers.  Data centers are just the last of a few of public service necessities which are desired, but don’t put them in my backyard.  Town halls, online sites, and social media are filled with irate people on both sides of the issue.

What is the big deal with data centers?  Well, broadband, big data, and a world of web applications has exploded the world wide web to be integrated into everything we do.  Broadband internet, web applications, wireless smart phones, fitness trackers, intelligent vehicles, smart homes, and smart televisions mean that every person is generating and consumering data in real time.  We don’t see data, but if it were bricks, there would be thousands of bricks streaming in and out of our houses, cars, and rooms every day.

The billionaire, as explained in my book America’s Twilight, is the beneficiary of IZATIONS.  Globalization, de-unionization, financialization, Walmartization, and robotization where all of these are technology advancements are used to make a single person wealthy by exploiting the date from thousands.  Instead, of the data making life inexpensive and enabling a higher standard of living, the wealthy billionaire uses data for one thing.  To get rich and to keep others from getting rich.  The Billionaire wants to extract money from global populations and redirect it to a concentrated few individuals in the world.

Data Center’s Nefarious Purpose

This is the point.  It is almost exclusively accomplished by having so much data flowing out of the world and into a few peoples’ hands, that the interaction from buyer to seller is completely one sided.  Big data centers are needed everywhere.  To track what Sally Jones buys at the store.  To see trends in buying habits of people in small-town, USA.  Understanding who likes today’s politics and who hates them.  To understand who is obese and who is eating too much chocolate.  Discovering service delivery and service quality before it becomes a problem.

True stories.

• An operations executive working can sit in California and discover every single union service technician’s substandard performance from across the United States and have a report on their desk the next day of the worst five of 400 technicians. They can see the ten longest installation appointments in a week or hour or day.  Highlight which installer is slow.  Sort out the worst five performers monthly and fire them for cause.  (De-Unionization)
• A big box store can determine what you want to buy before you enter. They can move their products and increase their prices before you enter the store.  In some cases, they even know you will pay more money for it.  They can increase prices of EVERYTHING.  A few pennies each, per million transactison, is a nice gain for one billioniare.  A few pennies increase across a store, times 100,000 stores, is a huge financial uptick.  (Walmartization)  The consumer won’t notice, right?
• A parking lot company can achieve better financial results by simply watching traffic in and out of the city. Higher traffic, higher parking fees.  Less traffic, lower parking fees.  They can increase prices from a single computer networked to 10,000 parking spaces (Financialization).
• A manufacturing company can now use digital twinning and global sourcing to set up and tear down manufacturing facilities in a matter of months. In doing so, the digital blueprint and the access to parts are so precise there is no need to ever have permanent manufacturing factories.  Simply move the virtual factory to Columbia … if that is where the customer sales orders are.  (Globalization)

The point is big data centers collect, manage, use, and distribute data in local satellites connected to the Billionaire “mothership in the sky” assuring the billionaire can continue making maximum money while we are on life support with the pennies afforded us.

That is the worst part of a data center.  The massive amount of data it aggregates is not in the best interest of any American citizen.

With that said, there are other more obvious elements of a data center that can be directly explained.

A data center is constructed with networking modules, computer data processing units, and data storage servers.  Every one of these data center systems and their associated equipment (circuit boards, power supplies, and cables) requires massive power and generates massive heat.

The numbers work like this.

1. A server/communications/processing printed circuit board might have 12 heat generating and power consuming components.
   1. The processing units mentioned above have ranges of different heat generating properties.
      1. Idle: 30°C – 45°C
      2. Average Load: 65°C – 85°C

• Maximum Limit: 90°C – 100°C (Throttle Point)

1. They generally consume somewhere around 170 watts to 360 watts each.

2. The circuit board is generally within an equipment chassis, which might contain four printed circuit boards. Now, 48 heat generating and power consuming components.
3. The equipment chassis might be 4 RU (U = 1.75 inches high times four Rus equals 7 inches) and is located in a rack. Racks are 19” to 24” wide and 36” to 50” deep.  They are often seven feet tall or 42 RU’s.  Now this means ten equipment chassis (40 RUs) are inside the rack.  This is now 480 heat generating and power consuming components.
4. A decent data center might have rows of racks. For the sake of argument, the data center has a row of forty racks in each row.  This is now 19,200 heat generating and power consuming components.
5. The data center has twelve aisles with two rows each. This is now 460,800 heat generating and power consuming components.
6. All equipment is typically designed with a heat sync. This is a specifically shaped solid metal plate that the components are married to.  The idea being the heat sync absorbes the component heat and moves it out.  Heat out meaning anywhere else.

Obviously, this would be enough heat to melt the printed circuit boards, the metal racks, the building, and everything in it.  What happens is that the powering and cooling strategies reduce the impact of the heat given off in these big data centers.  If the cooling were to fail, things start to get dangerous.

1. Equipment chassis are cooled by exhaust fans and chillers. These are basically a large industrial air conditioning system.  They can be installed in three ways.
   1. Most common, least effective – Cooling the room or the aisle and installing in or on an adjacent wall. They are much like the air conditioning units in your home.
   2. Average solution – Cooling the aisle or rack by directing cold air into the aisle or rack. Generally, from a duct system running under the floor or suspended from the ceiling.  And extracting hot air from the other side.
   3. Least common, most effective – Cooling the printed circuit boards of the equipment chassis itself by directing high pressure air into small ducts in every single chassis in the rack. For instance, in our example, every one of the ten chassis in a rack would have a duct in for cold air and a duct out for hot air.
   4. Where does this hot air go? It is vented out of the data center.  Therein lies the first problem … global warming.
2. Sometimes icy water flows into the equipment chassis and around the components to cool them down. Obviously, this might be a cold “heat transfer” liquid, instead of water.
3. In exotic situations, the heat generating and power consuming components are designed to work inside a tank, and the tank is filled with the cold “heat transfer” liquid.

In all of these scenarios, the basic flaw in the argument of data centers is three-fold:

1. Data centers power consuming components use enormous amounts of power 24/7. The power utility is based upon cyclical power.  For instance, at night, people go to bed and the power consumption of the city drops too almost nothing.  At noon, it might peak.  Data Centers do not behave this way but consume power all the time.  The rate based public utility model for cost per watt is distorted by this.  And as the utility is regulated to pass power costs to all consumers, the utility will simply pass this cost on to consumers in three ways.
   1. A massive increase in the power grids will distort peak power. It will last longer and become more expensive.  All consumers will lose savings in off peak hours and in peak hour premiums.
   2. Supply and Demand will dictate that there will be a massive increase in demand from one consumer, but will be viewed as an increase in the grid, the substation and the transmission lines. Therefore, increase prices for everyone.
   3. The increased steady consumption of power will drive the power utility to build more generations, transmission and distribution power equipment. In doing so, this capital expenditure will hit the rate calculation and justify increasing rates to recover the investment.
   4. Power supply and outages will reprioritize the need to sacrifice all customers for the needs of one customer … the data center. In this, the priority will be given to the consumer with the highest use of power.
2. Heat is ugly. With the data center that can approach 180 degrees Fahrenheit, this leads to all sorts of problems.  It can melt materials, be a safety hazard, and destroy data.
   1. It can vent heat, moisture, pollution, PFAS (Per- and polyfluoroalkyl substances), and other externally vented materials. This can leak water, fluids, air, dust, or moisture as the data center heats hot spots and vents hot hair.
   2. Heat expands, turns solids into melted messes, and turns coolants into vapor and boils water. Heat creates hazards that include melting dangerous plastics and composites.
   3. Even still, all of this can be transmitted into the environment by people walking in and out of facilities. It can be transmitted by vehicles.  It can be transmitted as damaged or obsolete materials and equipment are discarded.
3. Power, Heat, Pollution, Cooling. It is a vicious cycle.  And as each comes with a problem.  Data centers with more equipment than cooling end up being too hot.  Data centers with too little power cannot function.  Overpowering data centers with a power utility is IMPOSSIBLE to prevent other citizens’ power bills from increasing.  And Overpowering data centers using private power generation is dangerous, as they have no idea what they are doing.  Not like the experts at a power utility.

This is the snapshot of why data centers are problems.  The IZATIONS or the reasons these data centers are needed in the first place is often the most ignored and most important threat to the United States of America’s citizens.  Billionaires seeing into their homes and lives and leveraging this.

Power, Heat, Pollution, and Cooling are all problems.  But these can be solved.  The problem is that when looking at data centers already in place, the evidence suggests that the companies swearing they will be great neighbors have 10s if not 100s of data centers already in place that are not performing at this level of quality and safety.  How can you trust the people who are already doing a bad job at data center operation?