Options for exergy-rational use of heat generated by CHP systems

Overview

Birol Kılkış's paper at the 15th National Building Services Engineering Congress of the Chamber of Mechanical Engineers (TESKON 2023, Izmir) appears on page 357 of the proceedings and is freely readable. It develops an exergy-based optimisation model for using the heat generated by combined heat and power (CHP) systems. Its subject is bringing a thermodynamic viewpoint, one that looks at the quality of energy sources as well as their quantity, into data centre design.

The study scores five configurations against ten exergy-based criteria. One scenario is a near-zero-carbon data centre whose power and cooling plant burns natural gas blended with hydrogen and biogas (Scenario 2, NCL = 6). Other scenarios combine heat pumps, ORC generators and absorption/adsorption chillers. The conclusion is that CHP heat is best used as heat when demand exists, but for cooling a data centre the proposed cascaded combinations are the most rational option.

The author's central argument is that PUE is blind to the thermodynamics of such plants. In the innovative data centre scenario, the PUE of the trigeneration plant is worked out as (100 kW IT + 100 kW / cooling COP of 3) / 100 kW = 1.33, against a quoted sector average of 1.6. Yet PUE cannot distinguish the quality of the energy sources. First-law primary energy indicators (PEI) are also judged insufficient, and an exergy-based RUEEX metric is proposed to find minimum-CO2 fuel-to-cooling paths.

The study's limits are that the scenarios are computational and conceptual and the record contains no detailed measurement data; the PUE example is also a simplified calculation. The related standard is ISO/IEC 30134-2. The message for designers is that where on-site CHP or trigeneration is present, PUE alone can mask the choice of fuel.

Key points

1. Five CHP heat-use scenarios were scored with ten exergy-based criteria; Scenario 2 is a near-zero-carbon data centre (NCL = 6).
2. Example PUE of the trigeneration DC: (100 kW IT + 100 kW / COPc of 3) / 100 kW = 1.33, versus a quoted sector average of 1.6.
3. PUE is judged inadequate to describe the thermodynamics of on-site CHP/trigeneration DC plants.
4. First-law primary energy indicators (PEI) are also insufficient; an exergy-based RUEEX metric is proposed to find minimum-CO2 fuel-to-cooling paths.
5. Conclusion: CHP heat is best used as heat when demand exists, but for cooling a data centre the proposed cascaded combinations are the most rational option.
6. Other scenarios combine heat pumps, ORC generators and absorption/adsorption chillers.
7. Related standard: ISO/IEC 30134-2.

Using it in design

• In the energy module, when on-site CHP/trigeneration is modelled, report PUE together with source-aware metrics (primary energy, CO2 and optionally exergy-based figures).
• Warn that PUE alone can mask fuel choice.
• Catalogue absorption/adsorption chillers with thermal COP and driving-temperature fields.
• Check first whether there is demand for the CHP heat; where there is, prefer using it as heat over cooling.

Who should read it

Mechanical and energy engineers and academics designing data centre energy plants with CHP or trigeneration systems.

Related standards

• ISO/IEC 30134-2

Publication details

Authors

Birol Kılkış

Publisher

TMMOB Makina Mühendisleri Odası (MMO)

Year

2023

Document number

TESKON 2023 Bildiriler Kitabı, s. 357

Event

15. Ulusal Tesisat Mühendisliği Kongresi (TESKON 2023), İzmir

Document language

English

Source type

Industry body

Access

Free to read

Licence

Copyright retained; citation only — the publisher’s terms apply to the original document

Citation