paul evenson thursday, december 4, 2003 icetop tank manufacturing paul evenson university of...

Paul EvensonThursday, December 4, 2003

IceTop Tank Manufacturing

Paul Evenson

University of Delaware

December 4, 2003


Objectives

• Primary: Produce blocks of clear ice approximately two meters in diameter by one meter deep. Each block is to be viewed by two optical detectors (DOM), which are to be “frozen in” to the ice.

• Secondary: – Bottom and sides of the block of ice must be covered with a

diffuse, highly reflective material. – Entire assembly must be light tight.– The entire assembly must be insulated to an R value of

TBD.• Minimize the amplitude and suddenness of temperature

variations • Limit cracking of the ice• Meet environmental constraints of the DOM.


Approach

• Based on natural ice growth on lakes.• Clear ice is produced by a method known in the

materials industry as “zone refining” which exploits the tendency of the crystal forming from the liquid phase to exclude impurities that concentrate in the remaining liquid.

• In a lake, the “impurities” (which include the oxygen fish need to survive) are diluted in the large volume of lake water under the ice.

• Technical issues to be faced at Pole all derive from the need to conduct the freeze in a volume of water comparable to that of the end product.


Design Evolution: Major Tests

• 2000-2001 at Pole

• 2001-2002 at Pole

• 2002-2003 at Wilmington

• 2003-2004 at Pole


2000-2001 at Pole

• Pressure was relieved by means of a heated rod that was slowly raised as water froze in from the sides and bottom

• A large crack developed from the poorly relieved pressure.

• Optical signals were large enough to be useful.


2001-2002 at Pole

• Demonstrated a workable pressure relief system.

• Took 28 days to freeze.• Pressure relief system intruded

into the prime optical space of the tank.

• Quality of the ice was better, with no large cracks

• Ice still quite cloudy due to the inclusion of gases dissolved in the water prior to freezing.

• Once a plug of ice forms in the top of the tank, it cuts off any route for dissolved gases to escape to the atmosphere.


New Conceptual Design

• Pressure relief route was changed from a tube inserted from the top of the tank to a pipe entering the tank at the bottom.

• Water was continuously withdrawn at the bottom of the tank, run through a degassing device, and reintroduced at the bottom of the tank.

• Small scale tests demonstrated that essentially perfect blocks of ice could be produced in this way.


Degassing Principle

There are three commonly used methods of degassing water: heating, mechanical agitation, and permeable membranes.

We elected to use the permeable membrane approach, primarily because it is the only one that has a completely closed circuit for the water being treated.


2003-4 Wilmington Freezer

• Primary issue: Is processing the water only at the bottom of a large tank sufficient to keep the concentration of dissolved air at the freezing front at an acceptable level?

• The diffusion rate of air in still water is far too slow to permit a freeze in a reasonable amount of time.

• We demonstrated that the use of thermal instability can achieve the required mixing, and thereby avoid the need for a large mixing pump, either inside or outside the tank.


2003-4 at Pole

• Successful deployment of two (pre) production prototypes

• Incorporate design changes from PDR

• Concepts are proven, the tanks are still freezing as of 4 December 2003


Tank Manufacturing

• Starting point– Concept and operation of all systems have been

demonstrated• Issues

– Efficient manufacturing of 160 tanks– Efficient deployment of 160 tanks– Reliability and/or redundancy vs cost– Monitoring systems vs cost

• Next Goal– Deploy eight production prototypes during the

2004-2005 season


Production Tank Design


Major Components

• Insulated Tank• Insulating Pallet• Top and Lid• Degassing System• Pressure Relief System• Freeze Controller and

Monitor• Power System• Sunshade• Water Supply• Repair System


Insulated Tank

• Spin molded black polyethylene

• Polyurethane spray-on insulation

• Tyvek reflective liner

• We are investigating using an integral reflective coating on the interior of the tank to reduce assembly time and improve the quality of the fiducial volume


Insulating Pallet

• Light plywood or molded polyethylene form

• Filled with polyurethane spray-on insulation

• Slots for fork lift access• Current units use a standard

wooden pallet with insulation on top. This is much heavier than it needs to be since the full weight of the filled tank must be transferred first though the insulation and then through the pallet.


Top and Lid

• Must support the weight of two people or a snowmobile

• Frame construction with removable panels for cooling during freeze

• Supports freeze controller

• Supports DOMs during freeze-in

• Mounting for sunshade


Degassing System


Pressure Relief System


Pressure Relief System

• We are planning a significant redesign of this system to make it easier to install

• The degasser components will all be housed in a sealed assembly below the tank

• The new design will allow a definitive leak test prior to filling the tank

• Intrusions in the fiducial volume will be minimized

• Sump assembly will be re-usable to save on cost


Freeze Controller and Monitor


Freeze Monitor Program


Redundancy and Monitoring vs Cost

• The vacuum pump is the only required external active component

• Much of the present system consists of performance monitoring and backup control


Power System

• Requirement is 200 watts continuous, 800 watts peak

• Designed to survive an eight hour power outage

• System now at Pole uses a commercial UPS to provide both backup power and peak power for the vacuum pump (800 watts with near zero duty cycle)

• Other designs are possible but an approach consistent with the cable design needs to be frozen soon


Sunshade

• Based on “Bimini” style boat top

• Also serves as work shelter during assembly of freeze controller


Water Supply

• Standard station water is suitable

• Use of drill water may require purification

• Impurities precipitate out, degrading the optical quality of the bottom liner of the tank

• With knowledge of water chemistry details, the effect can be calculated and modeled

• Due to the different absorptive properties of different minerals, no blanket statements may be made


Repair System

• By injecting warm water through a tube in the relief pipe, and recycling the overflow via the sump pump, we have demonstrated that we can melt back the ice to repair internal defects

• The injection tube is installed in the tanks now at Pole, but to this point we have no specific plans to produce the required external units

• It is probably cheaper to abandon the tank and replace it with a new one than to try to salvage a defective tank

• Techniques for salvaging the DOMs in this case need to be developed


Production Strategy

• Insulated Tank– Procure from vendor – Final assembly and foaming at, and

shipment from the tank vendor’s facility– Continuous supervision by University of

Delaware personnel• Insulating Pallet

– Procure from vendor– Deliver to tank vendor for integration

• Top and Lid– Procure from vendor– Deliver to tank vendor for integration

• Degassing System– Assemble and test at University of

Delaware– Ship to tank vendor for integration

• Water Supply (TBD) • Repair System (TBD)

• Pressure Relief System– Assemble and test at University of

Delaware– Ship main relief system to tank vendor for

integration– Ship sump system to Pole for field

installation on tank (re-useable)• Freeze Controller and Monitor

– Assemble and test at University of Delaware

– Ship to Pole for field installation on tank (re-usable)

• Power System– Procure from vendor– Ship to Pole for field installation on tank

(re-useable)• Sunshade

– Procure from vendor– Ship to Pole for field installation on tank

(re-useable)