It is long overdue but I congratulate the forum committee for creating this forum topic as a standalone item. My feeling for a long time has been that energy efficiency and building performance is the key to long term success in
the 'controls' industry. I'm looking forward to this forum really exploding with discussion over the next little while. My open ended question to try and promote some discussion is a simple one. Analytics is a red hot topic
right now. My opinion is this is deservedly so but right now is the hype living up to the reality and is there a risk that building owners will be underwhelmed if analytics is not implemented well? Thoughts?
I think it already has not been implemented well. I believe it can be though. I'm not even really sure what IT is...is is proprietary knowledge that someone has transfer to code that gets used over and over? Is anyone sharing what they are actually doing down in the nitty gritty? The only things I have found so far online are APAR and VPACC.
1. The greatest problem facing building efficiency is that no one really wants to pay for it. Cheap wins big in the marketplace.
2. The second problem is building trades charlatans posing as experts.
3. Regulation as a solution is a pipe dream. Existing building codes are about safety and design gets lip service. Inspectors may be stupid or corrupt. Etc
4. Finding problems after creation is too late.
5. Even professional engineers screw up.
Consider oversized equipment, undersized ductwork, and leaking ductwork. These energy wasting practices in the HVAC trade are well-known but continue unabated.
Bottom Line: Human nature and ignorance corrupts everything.
It is also important to understand that for controls to accomplish the design goal they need accurate input. As Lynn stated on the cheap is dominate. For example there are water flow meters that will not accurately track variable water flow but the reps will swear they do and they end up being installed. Air flow stations need the calibration checked but some reps swear they are accurate in all applications from the factory. A balance contractor that takes the time to give the owner what they pay for has little chance of landing a big job.
Some engineers even after having a mistake proven to them will make the same mistake repeatedly and never admit they were wrong. They have little or no understanding of how to successfully test an installed system and often put the balancer in the position of getting the best results possible as installed.
Correcting the problems starts with the owner but they hire others because they admit they don't know but are often economically forced to put constraints on the design. Analytics are great but cannot compensate for everything.
One of our local high schools,about 25 years ago, invested big money in a campus wide chilled water and ice bank storage system. From the start, the occupants complained about the miserable comfort results. Maintenance, installing contractors, manufacturers, controls people, engineers, service technicians, occupants, administration blamed one another bitterly. Nothing helped. Nobody could be proven to be responsible. Finally, after 10 years, the school district quietly scrapped their investment. Taxpayers lost again.
"I have never let my schooling interfere with my education."
Mark Twain
NEVER STOP LEARNING.
Lynn,
If the balancer did his job he should have known the why and should have had the problem/s in the balance report. The balancer could very well have been intimidated into keeping quiet. I did a school that had 10 roof top ERV's with no source of heat recovery. Even the toilet exhaust did not exit thru any of the ERV's. I informed the mechanical I was under contract to and they forwarded my concerns to the general. After we met for over an hour the general said "frankly I don't see a design problem". My response was "what you see or don't see doesn't have anything to do with my report and I have no choice but to identify design errors". He responded "I hope you don't do that". I did it but nobody paid any attention to my comments and to this day the owner has the false belief they are saving energy.
At a different school, I requested to view the documents submitted by the installing AC contracted. On review, I found that the balancing " report" was, in fact, a detailed proposal to balance the chilled water system (water and air). It never happened. The documents were simply filled away. The school and the balancing company were screwed. The mechanical contractor pocketed a tidy bonus. THE END
.
Last edited by lynn comstock; 05-22-2017 at 05:43 PM.
trying to get Analytics back on topic -
I think most FM companies now have skilled (people in the main) in the positions that matter and energy management does carry a larger dynamic than it has ever had within the FM companies.
The analytics i think have to be structured so that they don't become an emperor's new clothes sales tool.
the analytics should by design give the customer what they need as part of a report, i doubt any customer needs to see a months worth of half hour data - they want to know what's using the energy and what can be done to reduce it if anything.
I think most of the problem with energy management is that in most cases it's always reactive, so when the energy manager or
I think most of the problem with energy management is that in most cases it's always reactive, so when the energy manager or customer gets a report at the end of the month the damage has already been done then an investigation has to take place as to why it happened and what can be done to stop it happening again.
I think analytic has to be incorporated into the BeMS (lowercase 'e' was intended Building energy Management Systems rarely are used to actually manage energy) at a grass roots level if an energy spike occurs during and 1/2 hours period that needs top be feed back into the BEMS and the BEMS can decide within a given set of parameters what it can do to minimise the impact of the usage spike.
I feel if analytics is used as another sales tool i.e. selling the information the customer already has back to him in another format, then it could quickly end up with a poor reputation and not be taken up as it should . Analytics has been around for years every BEMS has been carrying out those tasks since controls began - any PID control is analytics at work variable - setpoint - error - action - error reduction.
just my 2 pennies worth - i'm a BEMS engineer and have been saying all of the above for the last 15 years at least.
Whether or not analytics is used for improve climate control or energy-savings, it seems that good, well-designed alarms would be a more cost effective alternative to analytics. I actually search that topic: Analytics vs. Alarms - got sent to a SkyFoundry blog:
The DDC techs at a major US aircraft manufacturing facility, created "Energy-Maintenance" alarms to notify when the HVAC gear got into trouble - pretty effective. See attached PDF.
The skyfoundry article is correct in that to build an alarm you need to know the exactly what you're looking for. During my controls PM service days I found it good to sometimes to get away from the 'glass' (BAS display + 'alarms') and to just walk around. On day I came across this lonely AHU which made a 'thumping' noise. The unit was running and no 'alarms'. Knifed switch it off and took a peak at the motor - quickly found what the noise was about - a broken fan belt. The motor was running but the Veris CT was a simple DI type - current was flowing (in the old days they used analog CT for the motor status so you could see a current change - but more $$ for that CT). So a smart "broken fan belt" alarm might take a look at 1) If CT is True AND 2) coil delta T is below xxF (and the schedule is True) generate an alarm after so many minutes. If the fan is not generating a wind, then no heat is being taken out of the coil - so low delta T. Wonder if 'HAL' could figure out that out actually report 'possible broken fan belt at AHU-XX".
I agree that a lot of what is called analytics can and should be handled by the BMS with well designed alarm algorithms. Having said that I do think there is a place for analytics as BMS alarms, as you know, is primarily real time functions. I have worked with a couple of clients now with implementing a number of different analytics solutions. It seems to me that currently success in delivering real results is primarily a function of:
a) Having good quality engineers with access to quality supporting data (functional specifications, tech data etc) who have the time to investigate and resolve problems.
b) Implementing a good process to review, investigate, follow up and resolve issues raised.
c) Having involved and committed building owners and operations staff.
d) The capability of the analytics system, particularly with respect to having flexibility in developing algorithms. A lot of the out of the box stuff needs to be finessed particularly with complex HVAC plant.
In other words I think the human element is still key. I have done a lot of BMS upgrades / replacements over the years and more often then not the biggest benefit has simply been in having a knowledgeable engineer cast an eye over the control strategies and making sensible changes, usually the sort of changes that could be implemented on the old system.
Whether or not analytics is used for improve climate control or energy-savings, it seems that good, well-designed alarms would be a more cost effective alternative to analytics. I actually search that topic: Analytics vs. Alarms - got sent to a SkyFoundry blog:
The DDC techs at a major US aircraft manufacturing facility, created "Energy-Maintenance" alarms to notify when the HVAC gear got into trouble - pretty effective. See attached PDF.
The skyfoundry article is correct in that to build an alarm you need to know the exactly what you're looking for. During my controls PM service days I found it good to sometimes to get away from the 'glass' (BAS display + 'alarms') and to just walk around. On day I came across this lonely AHU which made a 'thumping' noise. The unit was running and no 'alarms'. Knifed switch it off and took a peak at the motor - quickly found what the noise was about - a broken fan belt. The motor was running but the Veris CT was a simple DI type - current was flowing (in the old days they used analog CT for the motor status so you could see a current change - but more $$ for that CT). So a smart "broken fan belt" alarm might take a look at 1) If CT is True AND 2) coil delta T is below xxF (and the schedule is True) generate an alarm after so many minutes. If the fan is not generating a wind, then no heat is being taken out of the coil - so low delta T. Wonder if 'HAL' could figure out that out actually report 'possible broken fan belt at AHU-XX".
A maintenance agreement with a honest HVAC contractor would prevent this. A facility like this would have this in place already, so once again sounds like the human factor of a tech not caring to check a belt caused this whole scenario.
I would say it will help forgetting about the term "Analytics", it is a word that everyone is using but devil is in the details.
I think articulating the top 5-10 problems solved by historical and current data helps understand the value of this topic. The top 5-10 problems should generate value in one or more areas of labor, equipment and energy costs
Having said this, anyone familiar with a list of use cases around data and data science for our industry ?
Regular Guest
TealNa liked this post.
Originally Posted by tridiumtech
Reply