What Makes Medical Cannabis Facilities Unique

by Andrew L. Shakely, PE, LEED AP BD+C

In 1996, California became the first state to remove state-level penalties for the possession and use of medical marijuana for approved medical conditions. Since then, through a mixture of voter ballot measures and legislative actions, a total of 28 states and the District of Columbia have passed laws allowing for use of medical marijuana. In 2016 alone, five states approved medical marijuana. Each state has its own requirements for the growth and processing of medical marijuana, but in those states that require indoor cultivation facilities, the design of these facilities must address unique conditions.

Production Goals

The construction and operation of a medical cannabis cultivation facility represents a significant investment, and the operators’ goal is to be able to produce a consistent, high-quality product that not only can be replicated between harvest cycles, but also maximizes product yield. To do this, the facility must provide a highly controllable, energy-efficient environment that minimizes risk of product loss. So, what are the key factors that make the design of these facilities unique?

High Energy Consumption

While companies are developing and experimenting with LED lighting options for grow lights, the standard of the industry remains high pressure sodium lights. Most cultivators believe high pressure sodium lights have proven to be the best lighting to maximize product yield. The impact of having these high-density lighting systems in an indoor facility is that they both consume a lot of energy and also produce a lot of heat. Not only does this require the electrical service and distribution systems to be much larger than with conventional buildings, it also causes the HAVC systems to be upsized to handle the high heat load of the lights.

Additionally, the HVAC systems must also be able to provide dehumidification in the grow rooms when the lights are off. During growing periods, when the lights are on, the plants are absorbing water and using it for photosynthesis. When the lights are off, a significant portion of the absorbed water is aspirated from the plants and released back into the grow room environment. The HVAC system must be able to control this humidity release, even though it doesn’t need to be providing cooling because the lights are off. Conventional HVAC systems aren’t designed to provide this type of humidity control when cooling is not required.

To reduce ongoing energy costs from operating this type of facility, it is imperative that the HVAC system be chosen with energy efficiency in mind. The energy savings from features such as free cooling and heat recovery loops will quickly offset the upfront costs of installing these systems.

Risk Control

Risk control in medical marijuana facilities typically falls into two categories: (1) Community Concerns; (2) Security and Crop Protection.

Security

Security is almost always a primary concern of both states and communities in which these facilities are located. Security systems are typically a significant portion of the cost of the project. Typical systems include perimeter access control and video cameras for monitoring both exterior and interior areas of the facility. They also feature access control and logging of employee movement into and out of critical zones of the building. These systems must have back-up power from emergency generators and UPS batteries, in the event of a power outage.

Odor Control

Another concern to most communities is odor control. It is vital that a thoughtful method of odor control be developed to limit odor from escaping the facility. Carbon filtration, ionization, hydroxyl, and controlled exhaust systems are techniques used to avoid conflict within the communities surrounding a grow facility.

Crop Protection

Crop protection is the primary risk control concern of the cultivator and operator of the facility. A pest infestation or outbreak of mold or fungi can spread rapidly through a grow room and cause a substantial financial loss to a grower. Indoor grow facilities must be designed with compartmentalized grow rooms to contain an outbreak, if it occurs. The materials used for walls, floors, and ceiling construction also must not promote mold growth and must be easily cleanable.

As mentioned previously, the HVAC systems must be designed to control humidity. Controlling humidity spikes is critical to avoiding an environment that can lead to mold or fungi growth. Systems must also not be shared between grow rooms so cross contamination cannot occur.

Because mold and fungi can cause a significant loss, the HVAC systems are also often provided with back-up power from a generator. The extent of backup power provided can vary depending on the reliability of the power system for the region in which the grow facility is located.

Controllability and Data Gathering

Because maintaining a consistent environment is critical to the success of an indoor grow facility, these facilities often feature advance control systems that monitor and control lighting levels, space temperature, humidity conditions, and control the fertilization and irrigation systems for the grow rooms. These systems also allow cultivators to continually optimize their grow practices by analyzing continuous data of growing conditions. Close coordination of the controls for lighting, HVAC, and irrigation systems is critical to avoid delays and added costs during systems start-up.

Summary

Indoor medical marijuana grow facilities have unique needs that must be addressed by proper design to ensure that the facility infrastructure supports the mission of producing consistent, high-quality, high-yield product. While this post has highlighted some of the key concerns that must be addressed during the design of the facility, it isn’t a complete summary. Other issues that need to be considered include water conservation, employee hygiene, and waste management. For an indoor grow facility to operate efficiently, the design must be a close collaboration between the owner, cultivator, and the design architects and engineers.

Skylights in Manufacturing Plants Might Not be so Green Afterall

by F. Joshua Millman, AIA, CFM, LEED AP

One of our clients recently approached us with a request to add skylights to a project that was under design, a new warehouse building. They saw several benefits: the natural light would decrease the need for artificial light and the cost payback should be great. Furthermore, their workers would benefit from having natural light. It would be another sustainable feature for the facility.

As gratifying as it is to have a client who wants to create as green a facility as possible, with a positive return on the additional investment, occasionally we need to limit the enthusiasm when the sustainable solution is actually less sustainable than a conventional design. Skylights fit squarely into that category for the following reasons:

  • Not the best return: If the most cost-effective solution to warehouse lighting is a project goal, then high bay fluorescent lights and occupancy/vacancy sensors offer a far better return than relying on passive lighting from skylights. A single 4-lamp T5HO fixture running for 8 hours a day, 20 days a month costs roughly $3.50 monthly. The local climate calculations will suggest the actual number of hours each year that are sunny enough to allow artificial lighting to be turned off. If one skylight replaces one light fixture, for an average of 20 hours per week, that’s only $21 in savings per year. 
  • Total construction costs: The costs of a skylight will depend on whether it will be part of new construction (less expensive) or cut into existing construction (more expense). There is additional cost for rewiring light fixtures to automatically turn off or dim. Should the skylight be strong enough to support a person falling onto it, or are guard rails around the skylight a better option? Both options add more cost. And sometimes a screen is needed below the skylight to catch any debris that might occur from a shattering of the skylight.
  • Life cycle costs: Any calculations on cost savings due to skylights must include life cycle costs. The flashing around a skylight is a maintenance concern and possible operational risk if a leak occurs. As designers, we do all that we can to minimize the number of roof penetrations. When natural light is mandated, we often propose clerestory windows or light wells, employing the same principle as the iconic sawtooth roofs of a century ago. However, if the roof must have smoke vents, these are available as skylights. Even the maintenance costs associated with cleaning bird droppings need to be incorporated into the thought process.
  • Effectiveness: Given land cost premiums, warehouses today are being designed ever taller, approaching the 45’ maximum height in which an ESFR sprinkler system can be deployed. The result is tall pallet racking that creates the visual effect of canyons. Although skylights can be centered on forklift aisles, how much natural light actually reaches the floor to provide adequate lighting to read markings or reach exits?
  • Actual time in warehouse: One key to effective warehouse operation is to minimize the manhours needed for operations. Short of full automation (AS/AR), the goal is typically to spend as little time in the warehouse as possible; complex algorithms are developed with the intent of reducing time in stacking and retrieving. Thus, the actual time people might benefit from being in natural light is to be minimized, decreasing any indirect benefits of skylighting.

Production Spaces

Do some of the same arguments apply to production areas? These areas typically have lower ceilings and are continually occupied. And yet, they can be plagued with additional unintended consequences resulting from skylights.

An example comes from a Maryland-based electronics manufacturing client. Our first tour of their plant revealed the aesthetic of tarps hung below the skylights, over the work benches from October to March each year. Our first thought was potential leakage from the skylight at its flashing to the roof, but there was no evidence of that from our inspection. We then concluded that conditioned/humidified air from inside the building was condensing on portions of the skylight installation. But how could that be?

The skylight vendor identified the insulation value of the skylight assembly as R=2.0. This is equivalent to a triple glazed window assembly. However, this is the average value; there were portions of the assembly where there was little or no thermal break. This design, coupled with the need to keep the plant at a relative humidity of about 40%, was calculated to determine that condensation would begin to occur at an exterior temperature of about +44oF. Such lower temperatures typically occur at least part of day in this region between October and March. Whereas interior condensation on a window may collect on the window sill, on a skylight it becomes dripping water. Some designs offer a way of collecting condensation in expectation that it will evaporate over time; other times the condensation exceeds the holding capacity. In the case of this facility, the condensation often froze at night, and then melted as the daytime became warmer. Janitors emptied the tarps of water each evening.

Another example is a Midwest bakery with skylights that experienced dripping only in the washdown area. For the same reasons stated previously, the humidity in that area resulted in an unacceptable amount of dripping from the skylights.

Ultimately, even minimal humidity, coupled with the coldest night conditions, will result in condensation in any building in almost any climate. The question becomes what amount of condensation, if any, is acceptable. In an electronics plant, some dripping a few days per year may be acceptable. In a food processing plant, condensation may be an opportunity for bacteria growth and insect infestation, no matter how well that condensation is contained within the skylight assembly, which offers another reason why skylights may not be an appropriate option.

The electronics manufacturer is now testing our proposed fix, which will reduce the dripping to +3oF exterior conditions. As for the bakery, they removed the skylights from the washdown area.

The Pitfalls of Designing too Precisely for the Future

by F. Joshua Millman, AIA, CFM, LEED AP

Highly unusual these days, it is to have a prospective client arrive in your office in a three-piece suit, even when it is not 83o outside.  Jared was a referral from a past client, who had described him as a “fellow facility manager”.  That was not my first impression as I met him in our lobby.  Most facility managers I know (and architects) have adopted “business casual” as the uniform of the 21st Century.  His LinkedIn profile reported that Jared had come up through the ranks in the finance area, most recently in corporate financial planning.  That background heavily informed the building planning process we undertook in the next two weeks.

Two weeks?  When Jared set that interval at the start of that first meeting, I was careful to not scoff aloud.  I plan buildings for a living, after all, so moving that fast is not as much a challenge for me as it usually is for the client who sets that high expectation.  But I had also worked for Honeywell as a corporate architect for several years, so such family-time-robbing turnarounds were sadly routine.

Jared had come to our meeting well prepared.  On his iPad was a PowerPoint show he had recently presented to his senior management.  The first slide focused on financial considerations:  the project budget, cost per square foot assumptions, and the project ROI (return on investment).  The next few slides indicated how much office space, production and warehouse space were needed in ten months, in five years and in ten years.  He even had a target utilization rate in the office area; that is, the ratio of usable area to gross area.  All the numbers were well within industry standards.  Two weeks for planning began to seem achievable.

The slide that was titled “Contingency Planning,” however, was the most intriguing.  The requirements were:

  • Ability to subdivide office area into multiple tenants; need central lobby/central restroom core.
  • Capacity to add a fifth tier of rack storage.
  • Flexibility to accommodate storage of hazardous materials to support possible future manufacturing processes.

The first item was clearly incorporating an ownership exit strategy into the building design:  there could be multiple tenants, even tenants who had no need for warehouse/production space, or tenants with no need for office space.  The other two requirements addressed building capacity and flexibility.

After he presented that slide, I asked him to hold on for a minute.  The expression on his face suggested that he was anticipating my praise for his forethought, perhaps the reaction he had received from his senior management.  I surely disappointed him then when I admitted that I did not think we could meet those requirements within his budget.  More importantly, I did not think we should meet all of them.  Here’s what I went on to explain:

Demisable Office Spaces:  Years of working for office developers have taught me that an efficient single-tenant layout is typically not an efficient multi-tenant layout.  As Jared had intuited, the placement of the lobby and restroom core is often the key.  A floor plan that could work for both would require more square footage, so we would not meet the utilization goals, which meant we probably would not meet the project budget.  My experience also suggested other considerations:

  • If the amount of office space needs to shrink, it’s usually less expensive to shut down some office areas than the construction required to demise those areas for a subtenant. The mothballed area becomes a reserve for future growth.  Otherwise, imagine having to later buyout the subtenant and then paying to remove the demising wall and rewiring the area for voice and data?
  • Office spaces connected to warehouse/production spaces have little appeal to those who want office space only. That level of disinterest usually translates to lower rents, further diminishing any payback from demising.
  • Bathrooms are less expensive to move than the cost of the additional square footage needed for an inefficient building. Better to design with flexibility to easily relocate restrooms.

Fifth Tier Rack Storage:  While the cost is relatively low to set a roof five feet higher to accommodate a fifth tier of racking, the following unanticipated costs could lower the return on the investment:

  • Depending on what is being stored and in what containers and what kind of pallets, the fifth tier may generate a code requirement for in-rack sprinklers.
  • The ability for skylights and/or clerestory lights to diminish the need for artificial lighting is itself diminished by the height of the roof and the canyons created by 5 tier high racked storage.
  • If the products stored need a tight range of temperature (or the people who work in the area are to have comfortable environmental conditions), then the increased height will aggravate temperature and humidity stratification, and large, slow-rotating fans will be less effective.

Hazardous Material Storage:  The cost of accommodating hazardous materials within a building demands that other alternatives be fully considered, whether the need is now or may occur in the future:

  • Certain levels of hazardous storage require the ability to contain any spills, as well as the water from the sprinkler system that might be activated at the same time as the spill. This is typically addressed by depressing the floor or adding a dike around the containment area.  Any liquid volume that cannot be contained in this area would be siphoned or pumped to an exterior tank.  Access into this containment area would most probably be via ramps for both forklifts and people.  Dikes could be added later.  There is even a door that can be installed in a floor recess that pops up as a containment gate when sensing any liquid in the recess.
  • Hazardous storage areas also may require multiple egress doors to the exterior; these doors’ sills need to be at the top of the containment walls, with landings on both sides, a ramp on the interior side, and often stairs on the exterior.  Casting additional doors in tilt up concrete panels is a further expense.
  • It may be less expensive to have those hazardous items stored offsite by others, and delivered on site in quantities that can be stored in a special cabinet to meet the needs of a few days.

By now Jared had taken off his jacket and loosened his tie.  Once I had answered his questions about each of my concerns in detail, these were the scenarios I recommended be calculated for each contingency item:

Scenario One: Determine the costs of building now without any future provision (Cost A) and constructing the retrofit later (Cost B).  Jared called this the “base case.”

Scenario Two: For the retrofits, identify which items would be significantly less expensive to put in now than later.  For example, add underfloor sewer lines to a future restroom area as part of the original building construction.  Calculate the cost of the building with the items best to do with the original construction (Cost C) plus the cost of finishing the contingency items in the future (Cost D).  This would be the “contingency case.”

At this point, the decision rested on both the ability to predict probability of the future need for each contingency and the availability of construction capital.  When applying this approach to the contingency items that Jared had proposed, we determined the following:

  • Ability to subdivide office area into multiple tenants: As predicted, it would cost less money to later demolish the restrooms and create a central restroom core than to design the office area to work for both single or multiple tenants (A+B>C+D).  Selling the building “as is” also seemed to be a more plausible exit strategy than trying to find a combination of multiple tenants to rent the building.
  • Capacity to add a fifth tier of rack storage: These calculations confirmed a negligible net cost to provide clearance for a fifth tier (A+B<C+D).  Having that additional height would probably make the building easier to sell, if necessary, given that there was also now room to create a mezzanine above three rack tiers of storage.
  • Flexibility to accommodate storage of hazardous materials to support possible future manufacturing processes: Given the low probability of this need, the contingency plan was created to build this later using containment dike walls and a popup gate rather than lower the slab, even though it would be less expensive to depress the slab now (A+B>C+D).  Drains and a lateral were installed in the slab for future connection to a containment tank.  These plans and provisions also supported the exit strategy.

This was then all worked into a great series of slides, complete with net present value analysis, for our joint presentation to Jared’s senior management two weeks later.  I even wore a necktie.

Interested in discussing the future of your facility? Contact Josh Millman, vice president of Nutec Design Associates, at 717.434.1570 or email him at jmillman@nutecgroup.com. 

What are You Doing with all that Data?

by Andrew L. Shakely, PE, LEED AP BD+C

Today manufacturers are swimming in data: inventory, WIP, output from each machine on the line, and of course, rejected parts and products.  But what is being be done with all this data?  Is it being used to identify where continuous improvement practices can be applied?  Are you even collecting the data that will have the greatest impact on your business?  The ability to amass mountains of data is relatively new, but Lean and Six Sigma practices have been around for decades.  So how can the data that is so readily available be used to minimize waste while enhancing the continuous improvement process?

Often, data within a manufacturing organization is compartmentalized within production cells or business units.  This data is typically used to maximize the performance of the individual unit, but the impact on the overall manufacturing process or organization may not be readily understood.  Focusing solely on out-of-compliance or defects alone within discrete units can create an atmosphere of finger-pointing and a defensive work force.  Instead, it can be much more effective to collect the diverse data and analyze it as a whole to look for the hidden interrelationships that are not apparent by simple compliance tracking.

The first step in this process is to develop methods to aggregate the data at single locations, and then dedicate resources to identify the underlying patterns in the information.  Use graphics and histograms to gain a visual understanding of the data.  What are the real moving averages and deviations?  Does the data reveal a pattern of disruption?   Once you understand the patterns within the data, you must look for correlations as well as cause-and-effects within the data.   Are there high correlations between events or is the data widely diverse?   Knowing these relationships can identify unexpected areas to concentrate your attention and help you determine which predictive indicators you really should be tracking.  Knowing the root-cause variables allows you to put controls and trending measures in place to improve output and reduce rejects.

It is important that you apply these analytics throughout your full production process, both upstream and downstream of where you think your “problems” are.   The hidden inefficiencies may actually not be occurring in processes within close proximity to each other, but rather from remote areas in your production.

The challenge to today’s manufacturers isn’t collecting the data, because it is already being collected everywhere. The real challenge is collecting and compiling this data coming from disparate sources, taking a methodical approach to analyze it, and understanding the patterns and correlations within the data.  With this information, you will have the knowledge to identify which production points should be the focus of attention, and where to apply Lean and Six Sigma practices that have will have the biggest impact on your organization.

Interested in discussing more about data? Contact Andy at 717.434.1526 or ashakely@nutecgroup.com