Latest / Mechanical Engineering Made Simple / (#25) The Engineering Secrets of HVAC: Why Cutting Fan Speed Tanks Energy Costs and the Physics of VAV Systems

Transcript

(#25) The Engineering Secrets of HVAC: Why Cutting Fan Speed Tanks Energy Costs and the Physics of VAV Systems · Mechanical Engineering Made Simple

• 0:00
  Welcome back, design jockeys, to Mechanical Engineering Made
• 0:02
  Simple. Today we are taking a really
• 0:05
  deep dive into the actual engineering behind controlling
• 0:08
  our indoor environments. We're going way past just, you
• 0:11
  know, fiddling with a thermostat.
• 0:12
  We're talking system architecture, thermodynamics,
• 0:15
  the real physics needed to keep a modern building comfortable
• 0:19
  all year long, no matter what's happening outside.
• 0:21
  That's exactly right. And today we're drawing on some
• 0:23
  core engineering fundamentals, design practices from pretty
• 0:27
  technical guidelines on indoor environmental control.
• 0:30
  We'll focus mostly on central HVAC systems, how you select the
• 0:34
  key components and that really complex job of air distribution,
• 0:38
  you know, duct design. OK.
• 0:40
  And before we jump into the deep end, here's our goal for you
• 0:42
  today. Stick with us and we're going to
• 0:45
  unpack maybe the biggest secret in HVAC efficiency.
• 0:49
  We'll show you why making a variable air volume fan a VAV
• 0:52
  system run at half speed doesn't just cut power use in half.
• 0:55
  Oh no, it absolutely tanks. It drastically reduces it, and
• 0:59
  that leads to, well, astonishing savings and operating costs for
• 1:02
  big buildings. It's a fundamental engineering
• 1:05
  trade off that honestly, it changed commercial design.
• 1:08
  All right, let's start unpacking this right at the core, the
• 1:10
  absolute heart of the operation, the central system doesn't
• 1:14
  matter if it's built up equipment, you know, constructed
• 1:16
  piece by piece on site or big sub assembled units shipped in.
• 1:21
  This is the engine room for year round conditioning.
• 1:23
  Yeah, the. Engine room.
• 1:25
  Good analogy. We're talking everything.
• 1:27
  The boiler for heat, the chiller for cooling, all the pumps, the
• 1:30
  piping and crucially, the air handling components.
• 1:33
  Those are the bits that actually touch the air going into the
• 1:36
  space. Right.
• 1:37
  And I guess the first big decision engineers face is how
• 1:40
  that conditioned air actually gets delivered.
• 1:42
  Exactly. And today we're focusing on what
• 1:45
  are called all air systems. That just means the system uses
• 1:49
  only air movement to handle both the sensible part, the
• 1:52
  temperature change, and the latent part, which is moisture
• 1:55
  control. And the type of system you
• 1:57
  choose dictates how much control you get in different areas or
• 2:00
  zones. Absolutely.
• 2:02
  The simplest setup is usually a single path system.
• 2:04
  These are pretty basic. All the air gets conditioned to
• 2:07
  one single temperature and then that same air goes through
• 2:11
  common ducts to every single area.
• 2:12
  OK, so. Cheap, but maybe not ideal if I
• 2:15
  have an office facing the sun and one facing north.
• 2:18
  They both get the exact same air temperature, yeah, which can be
• 2:22
  limiting. Right.
• 2:23
  Seems very limiting for complex places like, I don't know, a
• 2:26
  hospital or a university building where the heating and
• 2:29
  cooling needs are just all over the map, correct?
• 2:31
  So that must be where the dual path systems come in to play.
• 2:34
  You got it. Dual path is for hyper precise
• 2:36
  temperature delivery. You actually maintain 2
• 2:39
  completely separate air streams in ducts, 1 cold, one warm, and
• 2:43
  then these streams get mixed either right at the end just
• 2:46
  before the air enters the room. That's a true dual duct system.
• 2:49
  Or sometimes they're blended centrally back at the main fan
• 2:52
  unit. OK, more complex, but you get
• 2:55
  that fine control. Exactly.
• 2:56
  The complexity pays off because you can satisfy, you know, a
• 2:59
  freezing corner lecture hall and an overheating server room down
• 3:03
  the hall at the same time. OK, now let's talk about
• 3:05
  something that sounds, well, frankly nuts on the surface.
• 3:08
  The reheat system. Are you telling me engineers
• 3:11
  actually design systems that chill air down maybe to like 13°
• 3:15
  C essentially, only to then use more energy locally to heat it
• 3:20
  back up to maybe 18° C? That just sounds incredibly
• 3:23
  wasteful. Why?
• 3:24
  It does sound like a thermodynamic crime, doesn't it?
• 3:26
  But reheat is sometimes necessary.
• 3:28
  It's used when you need absolutely tight control over
• 3:32
  both temperature and humidity in a space, or maybe for perimeter
• 3:36
  zones that get wildly different sun exposure.
• 3:38
  So you often have to cool the air quite a bit just to wring
• 3:41
  the moisture out. That's the latent cooling part.
• 3:43
  And if doing that makes the space too cold, well, you have
• 3:45
  to add some sensible heat back in to keep people.
• 3:47
  Comfortable. OK, so the main goal there is
• 3:50
  humidity control 1st and then you fix the temperature
• 3:53
  afterward if you overshot. That's the idea, and to try and
• 3:56
  claw back some of that massive energy penalty.
• 4:00
  Good designs absolutely must use something called Reset control.
• 4:04
  This basically tells the system to keep the cold air coil
• 4:08
  running at the absolute warmest temperature it can while still
• 4:11
  meeting the cooling and dehumidification needs.
• 4:14
  It minimizes how much reheating you have to do later.
• 4:16
  OK, so reheat is the sort of necessary evil in some cases.
• 4:20
  Then the modern hero must be the variable air volume system VAV.
• 4:25
  Absolutely. VAV is just brilliant.
• 4:27
  It handles varying lows not by messing with temperature
• 4:29
  directly like reheat does, but simply by changing the volume of
• 4:33
  air supplied through a single duct.
• 4:35
  Room gets too warm? Open the damper, give it more
• 4:38
  cool air. Rooms getting chilly?
• 4:39
  Throttle that air flow way back. So each outlet, each diffuser
• 4:43
  basically becomes its own little controllable zone, Pretty much.
• 4:46
  Which means lower initial costs, right?
• 4:48
  Fewer ducts? Simpler controls maybe?
• 4:49
  Generally yes, single duct ones are cheaper.
• 4:52
  And the big one, the lowest operating costs of pretty much
• 4:55
  any all air system, especially in big buildings where one side
• 4:58
  is hot and the other's cold. That's the key advantage,
• 5:01
  especially with low diversity and VAV tech has gotten really
• 5:04
  good. Modern systems can throttle the
• 5:07
  air volume way down, sometimes as low as 10% of the maximum
• 5:10
  design flow. 10% Wow. Yeah, which makes it practical
• 5:13
  even for those tricky perimeter areas with huge swings and
• 5:17
  cooling needs. We should add, though VAV is
• 5:20
  primarily a cooling system at heart, it really shines in
• 5:22
  buildings with big internal loads.
• 5:25
  Think offices, data centers, places like that.
• 5:27
  OK, we've got the basic architectures.
• 5:29
  The what? Let's switch gears to the how
• 5:31
  much. If VAV is all about varying air
• 5:34
  volume, how does an engineer figure out the maximum volume
• 5:38
  they need to plan for? So we start with smaller
• 5:41
  systems, light commercial cooling stuff under say 10 tons
• 5:45
  or 35 kilowatts. OK.
• 5:46
  Yeah. For that kind of typical light
• 5:48
  commercial packaged unit, when you look at the manufacturer's
• 5:51
  data, you'll usually see a coil sensible heat factor, the SHF.
• 5:54
  It's typically around 0.75 to maybe .8.
• 5:58
  Sensible heat factor SHF. OK, what does that tell us?
• 6:01
  It's a critical number. It tells you what fraction of
• 6:04
  the units total cooling power is actually going towards lowering
• 6:08
  the air temperature. That's the sensible part.
• 6:10
  The rest is going towards removing moisture, the latent
• 6:13
  part. So an SHF of .8 means 80% of the
• 6:17
  cooling energy drops the temperature, 20% pulls out
• 6:20
  humidity. Exactly, and getting that
• 6:22
  balance right is crucial for comfort.
• 6:24
  People feel sticky if the latent load isn't handled, even if the
• 6:27
  temperature is OK. Now for just rough sizing,
• 6:30
  getting a ballpark air quantity for this kind of equipment,
• 6:34
  there is a common rule of thumb engineers use.
• 6:37
  The air quantity needed is typically about 400 CFN per ton
• 6:41
  of cooling. 400 cubic feet per minute per ton.
• 6:44
  Or in SI units, that's about 0.0537 cubic meters per second
• 6:50
  per kilowatt of cooling. You use that to get your total
• 6:52
  air flow, and then you divide that total amount among the
• 6:55
  different rooms based on how much cooling each specific room
• 6:58
  actually needs. OK, that sounds reasonably
• 7:01
  straightforward for a basic calculation like a snapshot in
• 7:03
  time, but you mentioned earlier that figuring out the real
• 7:07
  energy use over a whole year is actually way more complicated
• 7:11
  than just that rule of thumb. What makes it so tricky?
• 7:13
  It's the building itself. The biggest complicating factor
• 7:16
  by far is the thermal energy storage in the building
• 7:19
  structure. Think about it, The heavy
• 7:21
  concrete floors, the thick walls, even the furniture
• 7:24
  inside. They all soak up heat during the
• 7:26
  day like a giant thermal battery.
• 7:28
  And then they release it later. Exactly.
• 7:30
  They release it slowly hours later.
• 7:32
  So the sun might be strongest outside at, say, 2:00 PM hitting
• 7:35
  the walls, but the actual peak cooling load the HVAC system
• 7:39
  feels might not happen until 5:00 PM or even later as all
• 7:43
  that stored heat bleeds back out into the space.
• 7:46
  Precisely. You get this time lag, and the
• 7:49
  peak load the system sees can be quite different from the peak
• 7:51
  instantaneous load calculated from just weather and internal
• 7:55
  gains. So to get a truly reliable
• 7:58
  estimate of how much energy the building will use over a year,
• 8:01
  you really need a complex hour by hour computer simulations.
• 8:05
  These use typical weather data for the location and model how
• 8:08
  the building mass absorbs and releases heat.
• 8:11
  Simple hand calculations like for houses are really just
• 8:15
  adapted using average compressor runtime data from studies.
• 8:18
  You absolutely have to account for that building inertia.
• 8:21
  Got it. That thermal mass changes
• 8:22
  everything for accurate energy prediction.
• 8:25
  OK, so the building mass determines the load, but the
• 8:28
  fans are the muscles doing the work to meet that load.
• 8:32
  Let's talk about the air movers themselves.
• 8:33
  Right. The main workhorse, the one you
• 8:35
  see most often in HVAC, is the centrifugal fan.
• 8:38
  It's incredibly versatile. It can move small or huge
• 8:41
  amounts of air, and it can work against low or very high
• 8:43
  pressures, which makes it great for complex duct systems.
• 8:46
  OK, centrifugal, what's the alternative?
• 8:48
  Then you have axial flow fans, think venaxial or two back seal
• 8:52
  types. They look more like a propeller
• 8:54
  and a tube. These guys move massive volumes
• 8:57
  of air, but they generally can't push against high pressures as
• 8:59
  well, and they tend to be noisier.
• 9:02
  OK, different tools for different jobs.
• 9:06
  Now, when we talk about turning electricity into moving air,
• 9:09
  efficiency is key. How do we measure fan
• 9:12
  efficiency? There are two main ways.
• 9:14
  The most common one in standard US practice is total fan
• 9:17
  efficiency, which we write as ETA sub T Conceptually, it's
• 9:21
  just a ratio. How much useful air power did we
• 9:23
  get out compared to the electrical power we put into the
• 9:26
  fan motor shaft? And there's an equation for
• 9:28
  that. Yep, the standard form is ETA
• 9:30
  sub T = Q times the change in total pressure, which is P sub T
• 9:34
  1 -, P sub T2 all divided by 6350 times the shaft power W sub
• 9:40
  dish. Basically flow rate times
• 9:42
  pressure change divided by input power with a conversion
• 9:45
  constant. OK, so that lets you check if
• 9:47
  the fan is performing as advertised and you mentioned a
• 9:50
  second efficiency. Right.
• 9:51
  There's also static fan efficiency ETA subsets.
• 9:54
  It's calculated almost the same way, but it uses the change in
• 9:57
  static pressure instead of total pressure.
• 10:00
  A slightly different perspective on performance.
• 10:02
  Got it. OK, now for the really cool
• 10:03
  part. This is where the Vav magic
• 10:05
  happens. It's like fan loss.
• 10:08
  These are fundamental right. Absolutely fundamental.
• 10:10
  These govern how fans behave when you change their speed.
• 10:13
  There are three main laws that are incredibly useful.
• 10:16
  One, capacity. The volume of air moved your
• 10:20
  CFM, or cubic meters per second is directly proportional to the
• 10:23
  fan speed. Speed goes up 10%, flow goes up
• 10:26
  10%. Simple.
• 10:28
  Linear relationship makes sense to.
• 10:30
  The pressure the fan develops is proportional to the square of
• 10:33
  the fan speed. Speed goes up 10%, pressure goes
• 10:35
  up 21%. That's 1.1 ^2.
• 10:38
  Square law for pressure, OK. And three, this is the big one.
• 10:41
  The power needed to drive the fan is proportional to the cube
• 10:45
  of the fan speed, the. Cube right?
• 10:47
  So if you double the fan speed. You need 8 times the power 2 ^3.
• 10:50
  And conversely if you cut the fan speed in half like AVAV
• 10:53
  system does when the load drops. You don't use half the power,
• 10:56
  you use 1/2 ^3 of the power. Which is 1/8.
• 11:00
  You only use 12 1/2 percent the original power.
• 11:03
  Precisely that cubic relationship is the absolute
• 11:07
  game changer for VAV energy savings.
• 11:10
  Wow, so even though putting in the variable speed drives the
• 11:14
  VFD's or magnetic couplings costs more upfront?
• 11:19
  The drastically lower power consumption over the building's
• 11:22
  life just blows away that initial cost.
• 11:24
  That single piece of physics is why VAV dominates large
• 11:28
  commercial buildings. From an operating cost
• 11:30
  perspective, it's just huge savings.
• 11:32
  Amazing physics paying the bills.
• 11:34
  You could say that now, shifting from the powerhouse, the fan
• 11:37
  room, let's think about the delivery network, the ducts.
• 11:40
  Ultimately the goal of all this duct work, the air distribution
• 11:43
  system, is to create the right mix, temperature, humidity, air
• 11:47
  movement right there in the occupied space where people
• 11:49
  actually are. We need to achieve what we call
• 11:51
  acceptable thermal comfort. And just like energy is lost in
• 11:54
  wires, energy is lost in ducts due to friction, right?
• 11:57
  There's physics governing that too.
• 11:58
  Oh, absolutely. We use the total head
• 12:00
  relationship to understand it. The detailed equation is
• 12:03
  complex, but the core idea is simple.
• 12:05
  OK, give us the concept. The total pressure of the air,
• 12:08
  let's call it P sub O, Always, always decreases as the air
• 12:11
  flows down the duct. So if you measure it at .1 P sub
• 12:15
  O1 and then downstream at .2 P sub O2P, sub O1 will always be
• 12:20
  higher. The difference is the energy
• 12:22
  lost to friction, what we call lost head.
• 12:25
  Often written as I sub F. So mathematically something like
• 12:29
  P sub O1 divided by density rho, ignoring the G sub C factor for
• 12:34
  simplicity here equals P sub O2 divided by rho plus that lost
• 12:39
  head term I sub F. Essentially, yes.
• 12:42
  P sub O one over rho equals P sub O2 over rho plus I sub F
• 12:45
  Total pressure drops because of friction loss, and engineers
• 12:48
  spend a lot of time trying to minimize that.
• 12:49
  I sub F because every bit of pressure the fan has to generate
• 12:52
  to overcome that loss costs energy, costs money.
• 12:54
  Makes sense, and one way to minimize loss is choosing the
• 12:57
  duct shape, right? You often see rectangular ducts
• 12:59
  because they fit nicely above ceilings, but circles are better
• 13:02
  aerodynamically. That's correct.
• 13:04
  For the same cross-sectional area carrying the same amount of
• 13:07
  air, a rectangular duct will always have more friction loss
• 13:11
  than a perfectly round duct. It's just physics.
• 13:14
  More surface area for the air to drag against, relatively
• 13:17
  speaking. So how do designers account for
• 13:19
  that when they're sizing rectangular ducts?
• 13:21
  They use a concept called the equivalent diameter, usually
• 13:24
  written as D sub E. There's a specific formula for
• 13:27
  it. D sub E equals 1.3 * ( A * B
• 13:33
  close parenthesis raised to the power of 5/8, all divided by ( a
• 13:37
  + b ) raised to the power of 1/4, where A&B are the side
• 13:42
  lengths of the rectangle. OK, a bit complex, but the point
• 13:45
  is. The point is, it lets you
• 13:46
  calculate a circular duct diameter that would have the
• 13:49
  exact same friction loss per foot as your rectangular duct,
• 13:52
  so you can then use standard friction loss charts or software
• 13:55
  designed for circular ducts even when you're actually installing
• 13:59
  rectangular ones. It standardizes the calculation.
• 14:01
  Clever. OK, so once you know how to
• 14:03
  calculate friction, how do you actually lay out the whole duct
• 14:07
  system? What are the main design
• 14:08
  methods? Maybe for standard low velocity
• 14:11
  systems first? There are two main traditional
• 14:14
  approaches. The first is the equal friction
• 14:16
  method. It's pretty much what it sounds
• 14:17
  like. You design the system so that
• 14:20
  the pressure laws per foot of duct is the same everywhere for
• 14:24
  the main ducts on all the branches.
• 14:26
  OK, sounds equal. Simple.
• 14:29
  It's relatively simple to calculate, yes, that's its main
• 14:32
  advantage, but the downside is that the duct runs going to the
• 14:36
  nearest rooms will have very little total pressure loss
• 14:39
  compared to the runs going to the farthest rooms.
• 14:42
  So you end up needing to choke down the airflow to the closer
• 14:45
  rooms with dampers. Exactly.
• 14:47
  You often need significant dampering on the shorter runs to
• 14:50
  balance the airflow, and those dampers can be noisy if you have
• 14:53
  to close them down a lot. That's a major comfort issue.
• 14:55
  Right, Nobody wants a hissing diffuser overhead, so equal
• 14:59
  friction is maybe easy for the designer but potentially
• 15:03
  annoying for the occupant. What's the alternative?
• 15:06
  The alternative is often called the balanced capacity method,
• 15:10
  sometimes static regain, though that's more specific.
• 15:13
  The goal here is much more sophisticated.
• 15:15
  You try to size the ducts, such as the total pressure loss from
• 15:19
  the fan to every single air outlet is naturally equal or
• 15:22
  very close to equal. So ideally you wouldn't need
• 15:25
  much dampering at all. That's the goal.
• 15:27
  It should theory theoretically self balance the airflow to each
• 15:30
  outlet based purely on the duct sizing.
• 15:34
  It's much more accurate in terms of delivering the intended
• 15:36
  airflow, and usually results in a quieter system.
• 15:39
  But it's significantly more complex to design.
• 15:42
  It takes more calculation time, more iteration.
• 15:45
  OK, trade-offs as always. Now what about those situations
• 15:48
  where space is just incredibly tight?
• 15:50
  Like cramming ducks into the ceiling of a high rise building
• 15:54
  where every inch cost money. Right, when space is really at a
• 15:57
  premium, maybe due due to high construction costs or just
• 16:00
  architectural constraints, designers often shift to high
• 16:02
  velocity systems. Instead of large ducts with air
• 16:05
  moving slowly, you use smaller ducts with air moving much
• 16:08
  faster. Which means you need more
• 16:10
  pressure from the fan to overcome the higher friction.
• 16:13
  A lot more pressure, yes. You might see friction rates up
• 16:16
  to say 0.7 inches of water gauge per 100 feet of duct, which is
• 16:22
  quite high. But you can't just blast air
• 16:24
  into a room at that speed and pressure.
• 16:25
  It'd be like a hurricane. Exactly.
• 16:28
  So high velocity systems rely heavily on specialized terminal
• 16:31
  boxes located near the actual room or space.
• 16:35
  These boxes are crucial. They take that high pressure,
• 16:37
  high velocity air from the small duct.
• 16:39
  They throttle it down to reduce the pressure and velocity.
• 16:42
  They often have controls to regulate the flow volume like
• 16:45
  AVAV box. And very importantly, they
• 16:48
  contain sound attenuating materials to reduce the noise
• 16:52
  before the air finally enters the occupied space through a
• 16:55
  diffuser. So the terminal box is like a
• 16:57
  combination and pressure reducer, flow controller and
• 17:00
  muffler all in one unit near the room.
• 17:02
  That's a pretty good description.
• 17:03
  Yeah. OK, let's pull this all back
• 17:05
  together. We've gone through systems,
• 17:06
  sizing, fans, ducts, but the whole point, the entire goal of
• 17:10
  all this intricate mechanical engineering is ultimately about
• 17:13
  the people in the building. It's about human comfort.
• 17:16
  Right. Comfort isn't just temperature,
• 17:17
  is it? Not at all.
• 17:19
  It's a complex mix. It involves the air temperature,
• 17:21
  yes, but also the radiant temperature from surfaces around
• 17:25
  you, the humidity level, and how much the air is moving.
• 17:28
  Get any of those wrong and people feel uncomfortable.
• 17:31
  We actually have a standard for this, HHRA Standard 55, which
• 17:34
  tries to define the conditions under which at least 80% of
• 17:38
  occupants will find the environment thermally
• 17:40
  acceptable. That's kind of the benchmark.
• 17:42
  80% so you can't please everyone, but that's the target.
• 17:46
  And besides comfort, there's the health aspect, the actual air
• 17:49
  quality you have to bring in fresh air.
• 17:51
  Absolutely non negotiable. You need minimum outdoor air
• 17:54
  ventilation rates primarily to dilute the carbon dioxide that
• 17:58
  people breathe out, but also other indoor pollutants.
• 18:01
  The baseline minimum is generally never less than 15
• 18:04
  CFM. That's cubic feet per minute, or
• 18:06
  about 7.5 liters per second of fresh outdoor air per person.
• 18:10
  That's a fundamental safety requirement.
• 18:12
  And thinking bigger picture, this ties into environmental
• 18:15
  control generally, doesn't it? The HVAC system isn't just doing
• 18:18
  temperature, it's often acting as an air cleaner too.
• 18:21
  It really is, especially when you consider outdoor air
• 18:24
  quality. Many specialized air handling
• 18:26
  systems are designed specifically to deal with or at
• 18:29
  least mitigate common outdoor air pollutants, things
• 18:32
  classified as criteria pollutants.
• 18:33
  We're talking carbon monoxide, nitrogen dioxide, sulfur
• 18:36
  dioxide, ozone, and especially particulate matter.
• 18:39
  PM 2.5 and PM 10. The filtration choices, the air
• 18:42
  exchange rates, maintaining building pressure, these are all
• 18:45
  critical for protecting occupants from what might be
• 18:47
  outside. So the mechanical system becomes
• 18:49
  a key part of the building's defense against pollution.
• 18:53
  Fascinating overlap. OK, that brings us to our final
• 18:56
  provocative thought for you, our listeners, to chew on.
• 18:59
  We just spent time on how that cubic relationship in the fan
• 19:02
  laws makes VAV systems incredibly energy efficient over
• 19:05
  their lifetime because cutting speed saves so much power.
• 19:09
  So given that massive long term savings potential, why might a
• 19:12
  design team still sometimes choose a slightly noisier, maybe
• 19:16
  less efficient constant volume fan, like a simple centrifugal
• 19:19
  fan, if doing so significantly simplifies the ductwork layout,
• 19:23
  maybe uses less material initially, lowers the first cost
• 19:26
  of the project? Think about that inherent
• 19:28
  engineering tension, the trade off between spending less money
• 19:32
  upfront versus saving much more money over decades of operation.
• 19:35
  Where do you draw the line? Until the next equation design
• 19:37
  junkies, we are mechanical engineering made simple.