Monday, October 18, 2010

Energy Use: When should a separate power meter be required?

There's a fundamental flaw with many commercial leases in regard to power consumption - there are two separate matters that need protection, and the wording I frequently see protects neither.

The two issues are availability and consumption.  Consume more power than is available and you'll have a brown-out or a true outage, while higher than expected consumption kills utility budgets. 

The flaw is that the boilerplate text in many leases proposes a billing solution to an availability issue or vice versa. It's kind of like our government making a company pay a fine rather than helping that company fix the problem that led to the fine.

A little background
It's important to understand what is measured how.  Availability measures how much energy a building can provide at any moment in time, and is measured in kilowatts (kW).  Consumption measures how many units of energy a party consumes in a given period in time, and is measured in kilowatt hours (kWh). I tell people to think of it as a car lease... availability equates to the car's top speed, and consumption equates to the number of miles one travels in a month.
Think of it like a car lease.

It's also important to note that power clauses only appear in certain lease types. The leases we're talking about here are those in which the rent the tenant pays to the landlord includes power.  And not all such leases will contain a power clause because most tenants don't fall outside of a landlord's acceptable norms. I see this language more because technology clients tend to use lots of machines (go figure!).

Finally, note that for our puposes commercial power bills are primarily a function of (1) how many kWh you consume and (2) your peak instantaneous demand (kW) during the period. Having two variables complicates everything... the relative monetary importance of these two can swing in either direction.
And now for the challenge
Any time I've seen power addressed in a boilerplate lease, the landlord proposes to specify a kilowatt (kW) per rented square foot (SF) threshold.  If the tenant exceeds this kW per SF number, then the landlord can install a power meter on the tenant's space and can collect for power consumption overages. 

Going back to my car lease analogy, that's like saying that the if the driver exceeds 125 miles per hour then the leasing company can begin charging for consumption above the threshold.  Unfortunately, this approach leaves both the operations and billing concerns exposed:
  • Operations (availability).  The immediate effect of how fast I drive - a.k.a. my instantaneous kW usage - shows up as circuit overloads.  But the typical language doesn't address this issue by slowing me down or "beefing up" the engine... it imposes a billing-related solution.
  • Billing (consumption).  If we address how far I drive by installing a power meter, what is billed?  Am I charged for all miles driven above that speed and only those above that speed?  That likely won't do the landlord too many favors, as "peaks" are typically very short in duration.  Do you charge me for all excess consumption from that point forward?  Great... but in excess of what?  Consumption is measured in kilowatt hours (kWh), not kilowatts, so you can't charge me for consumption in excess of "X" kW/SF. 
If I'm laying this out right, it should seem pretty straight forward.  But it's not historically been done this way.  I recently asked a seasoned (and highly effective) real estate attorney with a well-known Atlanta firm if he's seen a kWh trigger in a lease before... he never had.  Then I asked my power and utility consultant connections about it and they were surprised to hear that leases address availability (kW) but not consumption (kWh). 

My experts did give me a caveat worth noting:  In data center environments, they do project kWh consumption, but they do so by assigning a utilization factor to the equipment's potential load to project kWh consumption.  In other words, they budget themselves to run at x% of peak all day every day, where X is a high number.  (Or in plain english, they guess based on historical data!) Perhaps the take-away here is that the leases I'm talking about should assign a load factor to equipment... but they don't.

The Solution
Rather than tracking potential loads of all installed equipment and negotiating a percentage of potential load to establish triggers for protect against outages and/or big power bills, I favor the idea of protecting operations (availability) and billing (consumption) separately.
Power Consumption = kWh

Availability: Find out what the tenant needs and/or what the building can offer (kW/SF) and determine in advance what will happen if utilization comes within x% of this level.  How do we determine how much of it is the tenant's fault?  If the tenant is at fault, what happens?  Will the tenant have to cap their consumption, install load distribution equipment, etc?  Will the building's capacity have to be increased?  Who will pay?

Consumption:  Determine the tenant's highest anticipated kWh consumption level and/or the building's kWh norms and agree to a base level for any excess billing.  In essence, agree that "If we go above this consumption level, then you can meter us and bill us for anything above it."

At the end of the day two factors will then be monitored... usage peaks (kW) AND consumption (kWh).  I recommend that they be monitored at the building level until the landlord becomes concerned, and then they drill down to the tenant level.  Operations is thus protected from service interruptions, and the landlord is protected from power bill increases.

An afterthought
It may be helpful if I expand a little on my biography here, because I do have some special insight into this topic. After getting my EE degree from Ga Tech, I spent four years designing power meters for Schlumberger's Electricity Management division.  I'm no longer an expert, but I certainly know a bunch of them!

I welcome your thoughts.

Here's to your success!


No comments: