Project Data

End-User: Lawrenceville School

Installer: Lighton Industries

Location: Lawrenceville, New Jersey

Date of install: May 2012

Size of system in kW: 6000kW

Energy produced since installation: ~27GWh

Special Challenges Met: Environmentally Protected Farming Land

Project Brief

THE LAWRENCEVILLE SCHOOL SOLAR PROJECT

Project Developer: TurtleEnergy LLC,  Date Built: 2012

Client: Sold to KDC Solar

Project: 6MW on thirty acres of single axis trackers, 500kW MV inverters

Cost of project: $28M, Date Energized: May 2012

John Millard's Role- principal in charge/ owner/ primary end user contact.

Civil, environmental, structural, mechanical consultant – Innovative Engineering; Wall, NJ

Permits required- NJDEP; Wetlands LOI, Flood Hazard Applicability Determination. NJ Department of Transportation- Permit to create curb cut on State Road 206. Township of Lawrence Zoning Board- Variance from Environmentally Protected Agricultural Zoning to Commercial Zoning Code. PSE&G interconnection permit, NJCEP SREC permit.

Role by John Millard: 1) Found TurtleEnergy 2003, 2) Sell Company Capabilities to Client in Nov. 2009, 3) Compete with two other companies to win $28M project, 4) Build project team to execute project, 5) Engage and Coordinate Engineering Consultants for Permits, 6) Negotiate Power Purchase Agreement, 7) Provide Due Diligence Design for Lender's Engineer to evaluate and validate for $15M non-recourse financing, 7) Estimate Annual Electricity Production for 20 year Production Guaranty, 8) Evaluate Mounting Hardware solutions for review, 9) Negotiate Procurement Contracts and Project Specific Warranties with Manufacturers of Inverters, Solar Modules, Mounting Hardware and Cable Tray. 10) Lead Site Layout and Subsurface Conduit Route for Permits, 11) Primary Liaison with Client, 12) Primary Legal Negotiator for Developer 13) Expert Witness at Zoning Board of Adjustment Hearing, 14) Negotiate SREC contracts with Load Serving Entities and SREC Brokers.

Project Achieved Zoning Approval in November of 2011 as well as all NJDEP Permits and NJDOT Permits- Project was sold to investors in April of 2011 and construction commenced in October of 2011, energizing in May of 2012 (commercial operations).

John Millard was the principal in charge of this $28M solar PV project. Millard introduced the concept of purchasing electricity from an on-site solar project to the school leadership and led the project development as well as the power purchase negotiations. Millard's company, TurtleEnergy, competed against two other utility scale project companies to win the 9000MWh / Year contract and prevailed.

Millard modelled the annual energy production that he forecast for the the project, communicated the intermittent behavior of the system to the utility and complied with the utility goal of constraining the AC output of the system to 5MW. This was accomplished by deploying a single axis tracker from Array Technologies who supplied the project with state of the art mounting hardware. The electrical design of the project, which was sited one half mile from the point of connection, used medium voltage (12.6kV) transformers to step up the native 200V inverter output in order to minimize line losses and to match the voltage that the utility metered power at the site. The school built a new load center for the campus that was coordinated with the solar electrical power requirements prior to the start of the solar project construction.The shovel ready 6MW solar project was then sold to a private power producer who provided the $28M fund to construct the project.

Schematic Design of The Lawrenceville Solar Project:

Millard formed TurtleEnergy in 2003 in response to the growing world demand for alternative energy solutions to a growing shortage of energy. Together with Millard’s architectural experience and a company called Turtle & Hughes, Inc., they became the nation’s first electrical distributor to form a renewable energy system integration, design & supply company. TurtlEnergy, LLC was founded by Millard and has grown sales of renewable energy systems from a $500,000 in 2004 to over $10,000,000 in 2006. Design and Supply of renewable energy systems is an activity which directly relates to the training which Millard received as an architect due to the highly multi-disciplinary nature of structural, mechanical, civil and electrical engineering coordination required in the renewable energy systems integration business. Hence, Millard formed a team of highly skilled architects & engineers with long histories in the renewable energy industry together with operational specialists within Turtle & Hughes. Prior to the Lawrenceville Project, TurtleEnergy had designed, supplied, construction managed and commissioned over forty solar projects in California, Massachusetts, Pennsylvania and New Jersey.

TurtleEnergy was a company created by John Millard in 2003 to design and supply photovoltaic energy systems to be retrofitted onto commercial buildings.  This activity took place within the constraints of the New Jersey Clean Energy Program and other state programs. This state program was conceived to create incentives for the construction of clean energy systems to produce electricity for businesses and homes in the state of New Jersey.

The Lawrenceville project with the culmination of eight years of business activity by TurtlEnergy which by then had designed over 30 solar electric systems for customers in California, New Jersey, Massachusetts, Pennsylvania, Connecticut and New York.

On-site generation of electricity by utility customers was a relatively new phenomenon on the East Coast. Most solar electric systems interconnected in New Jersey up until January 2010 we're limited to 2 MW AC by the Board of Public utilities in New Jersey. In February 2010, the Board of Public utilities removed the 2MW maximum system size in New Jersey. This regulatory change occurred in the middle of the process of awarding the Lawrenceville project to TurtlEnergy. The increased size of the system could now achieve 6 MW in order to satisfy the entire annual electricity budget of the Lawrenceville School which was 9,500MWh (or 9,500,000KWh- roughly $1M of electricity per year). Virtually tripling the size of the solar project overnight,      The Solar Energy Advancement and Fair Competition Act of 2010 turned our 2MW Solar Project into a 6MW project, an order of magnitude of increased change and economy of scale.

The increased economy of scale by tripling the size of the project not only reduced the unit cost of the project, but also enabled the sale price of electricity to Lawrenceville to be reduced by 40%. The cost of solar panels was still relatively high (compared to 2013 prices) but inverter costs reduced, installation prices reduced, and balance of system prices reduced (wire, mounting hardware, switches, tray, enclosures and other non-Pv components).

The reduction of these costs also helped to consolidate the preferred design option- a single axis tracker mounted PV system which rotated with the altitude of the sun over the course of the day (as opposed to the fixed tilt stationary mount that cost less). 

Schematic Design Issue: Mounting Hardware Selection

There are a number of advantages to the single-axis tracker despite its 5% increased cost to the overall project. Convincing others of this benefit, when they have concerns over the addition of moving parts to the mounting hardware is a challenge. This debate with parent company senior management was topical at the beginning of the project. Essentially, in order to produce the equivalent annual energy of a single axis tracker, one must purchase 15% more solar modules, inverters, mounting hardware and balance of systems to do so. This expenditure alone might silence a critic of moving parts, but there's more- that fixed tilt system redundancy costs to pay for. That 15% larger rated fixed tilt system is going to  produce far less electricity in the morning and late afternoon because of its azimuth angle facing due south. Since the fixed, south facing system makes up for this disadvantage at noon, by producing exponentially more power for an hour or two to make up for its poor morning and afternoon performance. This reality is overlooked when designing smaller systems because of the huge premium that a tracker adds to a small system such as a residential size. This spike of power at noon is accommodated by the utility with smaller systems, but not with Multi MW systems. The utility in the case of the Lawrenceville Project required a 5MW AC cap to the load profile of our design, largely due to the size of the utility wire that served the site. The larger, noon spike of the fixed tilt Pv system resulted in over 150 hours of +5MW operation that would have to be "clipped" or limited using current limitation devices at the point of common coupling in the newly constructed electrical building on the campus. The amount of electricity lost as a result of this clipping was significant enough to diminish any perceived advantage of using a fixed tilt system and the tracker was adopted at schematic design stage of the project. The failure mode analysis of the tracker's moving parts was peer reviewed by Black & Veatch and validated by those venerated engineers as investment grade. Our lender's due diligence engineer, Lahmeyer International also concurred that the tracker made by Array Technologies was "bankable" and the drive motor manufactured by Bonfigliotti was vastly under-employed in the tracker application. Turning the torque tube of mounted modules by one half of one degree every three minutes when it was designed to rotate many hundreds of RPM it seemed to them to be well within the specification of the motor. The slew ring that transferred this tiny rotation into the torque tubes of each module row did have a defective seal in early shipments to early projects in 2006, but this was a quality control problem in the Chinese assembly. All assembly is now in house at Array Technologies in Albuquerque. A site visit to the HQ prior to the $3M purchase of the 6MW tracker proved out the quality. 

Schematic Design Issue: Unforeseen Geotechnical Cost Mitigation:

Regardless of the tracker vs. fixed tilt mount, each of these systems were vulnerable to unforeseen geotechnical risks due to the base case cost assumption that these structures would find favorable ground conditions and the steel posts would be driven freely into the ground using a vibratory pier driver. This method of pier placement requires no concrete and relies on eight to twelve feet of penetration in order to support the lateral load of the wind driven module surface area. If the ground is a "hard dig" situation, it can drastically increase the mounting hardware cost compared to a self-ballasted solution such as Game Change's cast in place system. Our geotechnical study indicated a 20% chance of "pier refusals" which meant that additional equipment and/or concrete footings would be required to solve for the structural spec on each refusal. A pilot hole can be a lowest additional cost, where a drilling rig comes in behind the vibratory pile driver and reems the pier hole to depth and performs a pull out test or, if rocky condition prevail such that a pilot hole is insufficient for the task, then shallow concrete foundations may be the only solution. This is the worst case scenario and can add $100 to $200 per pier. These are the true risks of ground mounted solar, not the moving parts of a single axis tracker. 

Another challenge specific to the single axis tracker is the mid-module mount onto the 4" torque tube. This mount can result in bending the module around the midpoint and cause micro-cracking in the 150 micron thick silicon cells if heavy snowfall settles on the tracked modules while in stow (horizontal) position. Fortunately, thanks to the fact that the range of movement rotates the module every day over 90 degrees, it is impossible for snow to accumulate over more than a 12 hour period. Should 30psf accumulate during that 12 hours during the stowage overnight in horizontal position, then the bracketry that attaches the module to the torque tube must clamp at least 24" of the mid section of the module if the warranty on the module is to be preserved. Most module manufacturers archive an electro-fluoroscopic image of every module that they encapsulate and use this "x-ray" image as proof that the module had no cracks in the cells at shipment. In the case of the Lawrenceville Solar Project, 25,000 solar modules mid-mounted on torque tubes were sufficient cause to request a special project specific written warranty for re-assurance. This the module manufacturer did, and provided it with accompanying failure mode testing documentation of the solar module. The deflection of the module, and the resulting documented micro cracks were clearly defined so that tolerances were understood and engineered for in advance of installing the 6MW project.

Schematic Design Issue: Which is the greater risk? Electrical Risk or Mechanical Risk?

The mechanical opportunities for disaster on a large ground mount are by far more risk than the electrical risks, especially since National Electrical code section for PV has been so thoroughly forged. There have been ground mounted projects that were seamlessly designed and installed but for the near constant 30mph winds year round that induced a harmonic vibration akin to the "galloping girdie" effect that destroyed the Tacoma Narrows Bridge. The tracker manufacturer at Array Technologies had this mechanical risk in mind when he designed into the ends of each module row a hydraulic damper to quash any such movement that might emerge. The rest of the hardware involved with mounting the modules, the module clamps, the viscous couplings that capture the torque tubes, the slew rings, the drive assembly with the universal joints that enable a high degree of tolerance for unruly topography and other components are all well travelled designs from the automotive industry that have generations of predecessors. Even the mechanical component of trenching pipe & wire represents more risk than the electrical nature of the wire itself. Trenching electrical cable from a to b is an irresistible mode of operation for most electrical contractors but in practice, it causes nearly as much congestion and "hard dig" risk in practice and is labor intensive- pure profit for an EC and theoretically easy to estimate cost. Most Electrical Contractors do not automatically opt for cable trays as a substitute for trenching cable. In fact, ground mounted solar power projects are a natural application for cable tray. The long runs of mounting structures are easily suited to suspend tray from piers. Electrical contractors frequently trench right next to the arrays of peers that have been driven into the ground- only to be destabilized buy a 24 inch wide trench being cut inches away from the point where they were driven into the ground. Compaction of these trenches never matches the compaction that existed for that pier prior to a trench being excavated immediately next to it. Additionally, soil conservation districts are now requiring topsoil to be removed if extensive trenching is determined to be necessary for a given solar project in New Jersey. 6 to 12 inches of topsoil to be scraped off the surface of the ground and stockpiled on site-  silt fences and dust control measures have to be implemented by the contractor. Electrical contractors doggedly carry trenching as their primary distribution strategy for large ground mounted solar projects. At the Lawrenceville solar project, total energy specified cable trays to be clamped to the peers which for the linkage columns. These are the columns that have the driveshaft and the slew rings mounted on top. These are larger eight inch columns compared to the 4 inch piers that support the torque tubes of the solar panels. By mounting the cable tray to these columns, the vast majority of cable on the project can be transported above ground without having to excavate 36 inch deep trenches. Cable Tray, in lieu of trenches liberates movement of construction operatives across the site, unhindered by the hazard of open trenches, not to mention the impediment that erosion and mud can cause long term on any construction site. It was shocking to see how little knowledge there was of cable tray code within the electrical engineering discipline at the consultant level. When the owner expressed a preference for cable tray (because of the farmland that was going to be cycled out of production for the 20 year term of the power purchase agreement) our electrical engineer expressed that it would cost $1M to use tray comprehensively across the project (actually the material cost for the tray was $70k). The labor to install cable tray is less than the labor and equipment to excavate trenches (excluding topsoil stockpiling). Then there are all of the other benefits of cable tray- easy maintenance/ commissioning access, less copper per code, less cable, less loose dirt/ mud, and less disruption to free movement on site.