International Glazing Database – 

Verification Procedure for Glazing Data

1. Introduction

2. Acknowledgements

3. Purpose

4. Standards

5. Proficiency Testing

6. Choosing Representative Samples and Data

7. Submitting Data

8. Technical Review

9. Peer Review

10. Accelerated Review

11. Inclusion of Data

1          Introduction

The International Glazing Database (IGDB) is a publicly available collection of data for more than 1000 glazing products from manufacturers all over the world. Each record contains detailed spectral optical data, thermal data, structural details, description and product information.  The data is sufficient to design glazing systems and windows and perform accurate energy performance calculations.

2           Acknowledgements

Lawrence Berkeley National Laboratory (LBNL), with the support of the U.S. Department of Energy, performs technical review of data originating in the U.S. The WinDat Network, with support from the European Commission performs technical review of data originating in Europe.  LBNL also updates and maintains the database file itself. The database could not exist of course without the support of the glazing manufacturers themselves who measure and format the spectral data and provide additional product information. The manufacturers also play a critical role in peer review of all data.

3          Purpose

This verification procedure describes the necessary steps for determining, submitting, and reviewing data on glazing products for acceptance by International Glazing Database.  This data may then be used to calculate spectral averaged optical properties such as visible transmittance, solar transmittance or reflectance. These properties in turn can be used to calculate total window properties such as thermal conductance (U-factor) and solar heat gain coefficient (SHGC or g-factor).

4          Standards

4.1         Spectral Data

Raw spectral data shall be determined in accordance with standards published by recognized national and international organizations. For example, data originating in the U.S shall be collected in accordance with ASTM E903[1] and with ASTM E1585[2]. The choice of standards for this purpose is not particularly significant because standards of this type simply describe good operating practice for commercial spectrometers. Current standards from ISO, CEN and JIS are all acceptable.

4.2         Calculated Averages

Typical standards for measurement of spectral data also include procedures for calculating spectral averages from the raw data. In this case the derived quantities could be very different, depending on particular choice of weighting functions. It is a common misconception that this potential variance is of concern to us from the point of view of verification. The fact that the measurement practice and calculation methods tend to appear together is incidental in this case.
In addition to the spectral data, the IGDB does contain some nominal averaged values, but only for the purpose of viewing, sorting and filtering the entries. The standard by which spectral averages are ultimately determined is set in the software that reads the database. If a particular code or rating system requires that a certain standard be followed, that is a matter of software approval, not data verification.

5          Proficiency Testing

5.1         Frequency:

All submitters of optical data must successfully participate in proficiency testing by interlaboratory comparison every 4 years or sooner. The next interlaboratory comparison is scheduled to begin in April 2003, resources permitting.

5.2         Plan:

A detailed plan for interlaboratory comparison testing shall be developed in accordance with ASTM E1301.[3] 

5.3         Suspension:

Failure to meet the acceptance criteria of the proficiency-testing program may result in temporary suspension of rights to directly submit data. Every practical effort will be made to take corrective action. In the interim, measurement can continue at commercial laboratories that have met the acceptance criteria, in order to prevent delay in issuance of product labels.

5.4         New Submitters:

For new data submitters, the proficiency-testing plan shall include a provision for immediate participation in the program.  If the acceptance criteria are met, approval of the laboratory will be valid until the next testing cycle.

5.5         Report:

A final report of the interlaboratory comparison shall be published and freely distributed to all interested parties.

6          Choosing Representative Samples and Data

6.1         Production data

An important element of the data verification process concerns the choice of data that properly represents a given product. Optical properties may vary depending on many factors such as manufacturing site, the batch from a particular site, or uniformity of the particular batch. The U.S. glass manufacturers, for example, adhere to ASTM C 1036[4], which specifies a level of quality control. Some manufacturers perform extensive on-line and off-line diagnostics and maintain detailed records. Others rely more on inspection, experience, and less-frequent sampling.  The most rigorous solution is a statistical analysis of production data.

6.2         Quality control

Quality control issues in production are considered to be outside the scope of the IGDB verification procedure.

6.3         Sample selection methods

Specific methods for choosing samples are left to the individual manufacturer under general guidelines given in this section. Safeguards against biased choice of samples are built into the technical and peer-review processes described below.

6.4         Representative data

Data shall represent average solar and thermal properties for the product as sold by the supplier (prior to fabrication into IG units).  It is acceptable to submit measurements taken on many samples or a single measurement on a sample considered to be typical. Representative data shall be supplied for a nominal glass thickness, although actual samples can vary in thickness.

7          Submitting Data

7.1         Format

It is the responsibility of the data submitter to put the data into the proper format for review and subsequent merging into the database file. The format of the data along with examples for each type of glazing are specified in detail the IGDB Data Submission Procedures.[5]

7.2        Proprietary glazing information

Glazing data files must also include information about trade names, appearance and other apparent physical characteristics of the products. Glazing data files may include additional information about the structure, types of materials, and appearance of the glazing product.  No information will be required, however, about proprietary designs of coatings or formulation of bulk materials.

7.3         Information about calculated data

The derivation of data calculated from approved data components shall be clearly specified, including:

1. The sequence of the layers used to construct a new composite glazing such as a laminate or an applied film including material types and thickness.
2. The names of each of the approved files used to construct the new composite glazing.
3. The names of the approved files used to extract secondary properties such as the internal transmittance of PVB or the coating properties.
4. The program or algorithm used to perform the calculation.

7.5         Retention of samples

The submitter shall maintain the sample or samples used to determine the representative data for a period of two years. In the case of sensitive samples, such as those based on silver, every attempt should be made to protect the samples by storage in a dry environment. It is recognized however that deterioration may take place despite these precautions.

8          Technical Review

8.1         Format and Content

Initial review of format and content will be performed for each new file at LBNL. Data will be checked for conspicuous problems including discontinuity, excessive noise, nonphysical values, incorrect intervals and completeness of the file.

8.2         Self-Consistency

Comparisons will be made with other files in the same series or with similar products already in the database to check for consistency.

8.3         Duration:

Review shall normally occur within 1 week of submission. Delays due to foreseeable travel, vacation, or special projects shall be announced to the review groups as soon as possible. If problems are discovered with the format or the data, the technical review stage will be extended until the problems are resolved. 

9          Peer Review

9.1         Distribution:

Upon completion of the initial technical review, data shall be distributed for peer review. At this stage, representatives of major glazing manufacturers perform review. At minimum, the peer review group shall consist of all approved data submitters.

9.2         Duration

The peer-review phase shall last for 2 weeks. If no questions are raised during this time, then the data automatically passes.

9.3         Objections

If a question or objection is raised, the objector shall be required to provide the following information by email:

1.      The name of the product and data file in question.

2.      A clear statement of the suspected problem.

3.      Any supporting evidence for the objection. This evidence might include purely technical data, such as a measurement on a similar sample, or it might be nontechnical, such as a reference to contradictory product literature.

9.4         Resolution:

The data supplier shall be given the opportunity to resolve a misunderstanding or otherwise demonstrate that the objection is not valid. This process shall occur anonymously through the intermediary of the technical review team.   If the objector is not convinced, then one or more samples shall be requested from the supplier. If the results agree with the originally supplied data then the objection shall be overturned.

9.5         Correction and Withdrawal

If the measurements do not agree with the originally supplied data then the supplier can either withdraw the data from consideration or resubmit the data after taking corrective action.

10      Accelerated Review

An accelerated review may be performed for data under special circumstances. This accelerated review shall consist of simultaneous preliminary review and peer review. This accelerated review shall normally occur within 1 week. The following conditions must be met:

1. The submitter must specify by email that without rapid review a product release or building project would be delayed.
2. The technical review team must have the time available.
3. There must be no objections from any peer reviewer within 24 hours of  release of data.

11  Inclusion of Data in the Database

11.1     Updates and Posting:

Upon completion of the peer review term, LBNL shall as soon as possible inform all reviewers and other interested parties by email. Then LBNL shall prepare a database update program that will be posted on the IGDB website. Download and execution of this program will automatically add the new data records to the previous MS Access database file. Until April 1, 2003, LBNL will also prepare an update package for the text file collection used by WINDOW 4.1 and other programs.

11.2     NFRC Acceptance:

A special field in the IGDB designates NFRC acceptance of the data.  The value of this field is determined solely by NFRC, according to their internal requirements. If a data submitter specifies that they are seeking NFRC acceptance of their data, then NFRC will be informed upon the start of review.  If and only if NFRC confirms that the data has been accepted, will the field be activated.