1. Can a Window with a Higher U-Value Perform Better in the Energy Balance?

Can a window with a higher (less favourable) U-value appear better in an energy balance because it has a higher g-value? In a purely arithmetic balance, this is possible. Whether it actually reduces heating energy demand is another question.

A higher U-value means greater transmission heat loss. This is unambiguous—but it describes only the loss side of the equation. A window is not an opaque wall element. It loses heat to the outdoors while also admitting daylight and solar heat. What matters, therefore, is not only the annual net balance, but also whether solar energy enters during hours when there is a positive demand for heat.

This is where the apparent contradiction lies: windows are usually assessed according to how effectively they retain heat. Their second energy-related function—bringing solar energy into the building—is often treated merely as a boundary condition. Yet on a sunny south-facing façade, it can fundamentally change the balance, whereas its contribution on the north side is much smaller.

The U-value answers the question: How much heat does the window lose? The energy balance goes further: What remains after transmission losses, solar energy entering the room and the usable share of that energy are taken into account?

2. Why the U-Value Matters—but Tells Only Part of the Story

The U-value is indispensable. It describes the heat flow through a component per square metre and per kelvin of temperature difference. For a complete window, the Uw-value is normally used; it includes the glazing, frame and edge-of-glass assembly.

For instantaneous transmission power, the model uses: P_trans = Uw × A × (T_i − T_e). P is expressed in watts. Over a period of time, the corresponding energy is Q_trans = Σ(P_trans,h × Δt), expressed in Wh or kWh.

A simple example illustrates what this value tells us: a 2 m² window with Uw = 1.3 W/(m²K) loses 52 W at a temperature difference of 20 K. At that particular moment, the relationship is clear: the lower the Uw-value, the lower the transmission power.

However, the U-value does not tell us whether sunlight is falling on the window. It contains no information about orientation or shading by window reveals. It does not distinguish day from night or winter from summer. Nor does it indicate whether the incoming solar heat is needed in the room or instead creates a cooling load.

Thermal comfort may also be relevant, for example because of cold internal surfaces. The following analysis, however, focuses on the window’s energy balance.

3. Solar Gains and the g-Value

The g-value is the key parameter on the solar side. It describes the fraction of solar radiation incident on the glazing that enters the room as heat. This initially calculated solar energy input is not necessarily the same as a useful solar heat gain.

Determining this quantity correctly is not a minor issue. The scientific literature distinguishes standardised test procedures from newer measurement approaches for determining the solar heat gain coefficient, or g-value. The measurement methods themselves demonstrate that the solar performance of a window is a distinct building-physics property—not a decorative add-on to the U-value.

In the calculator, instantaneous incoming solar power is given by: P_solar,in = G_vertical × A_glass × g × F_shading × F_reveal. The energy accumulated over a period is likewise Q_solar,in = Σ(P_solar,in,h × Δt).

The frame reduces the effective glazing area. Window reveals, neighbouring buildings, vegetation and solar shading reduce incident radiation. Incoming solar energy becomes useful only when it reduces an existing heating demand in the same hour. If the room is already warm, that same input may remain unused or increase cooling demand in summer. The building’s actual heating demand also depends on ventilation, thermal mass, internal gains, controls and other building components.

For this reason, saying “a high g-value is good” is just as incomplete as saying “a low U-value is good”. Both statements are meaningful only in the appropriate context.

4. Orientation, Time of Day and Season Change the Assessment

The same window can play four different roles, depending on the circumstances:

The annual energy balance therefore does not arise from one steady-state condition. It is the sum of many hours with differing outdoor temperatures, solar irradiation, shading and usage conditions. Consequently, even identical windows can produce markedly different results on north- and south-facing façades.

Time resolution is thus critical to the assessment. An annual average can indicate the overall balance, but only daily and seasonal profiles show when gains occur, when losses dominate and whether a switchable operating state could be beneficial.

ISO/TR 52016-4 is referenced here solely as a methodological basis for time-dependent analysis of adaptive building-envelope components. The hourly window comparison model used here does not constitute a normative assessment under ISO 52016.

5. Illustrative Energy Balance of a Window

The following previously documented example from Calculator 2 compares the same older double-glazed window facing north and south. Only the orientation changes; all other inputs remain identical. This is a modelled component-level energy balance, not a measured whole-building heating demand.

Q_+ denotes the sum of the positive hourly window balances: Σ max(0, P_trans − P_solar,in) × Δt. This purely component-level calculated quantity is not actual building heating demand. Any different wording in the calculator interface does not alter this physical distinction.

Q_solar,in describes the total solar energy transmitted through the glazing. A useful solar gain is only the share that simultaneously reduces an existing heating demand in the room or building. Calculator 2 does not report this usable share as a separate normative indicator. Likewise, Q_− is a calculated surplus in the window balance, not automatically usable heat for space heating.

The example illustrates the central issue: Uw, area and the transmission balance are identical for north and south. Solar energy input changes substantially. Yet Q_+, the positive hourly window balance, decreases only from 177.9 to 168.9 kWh/m²a in the south-facing case, because much of the additional solar energy arrives during hours when the window no longer has a positive residual demand. This temporal distinction is precisely why the 511.8 kWh/m²a of incoming solar energy must not be mistaken for an equally large reduction in heating energy demand.

Second Comparison: Different Uw- and g-Values

Now the location (Bielefeld), south-facing orientation, 1 m² window area, 20 °C indoor temperature, 25% frame proportion, shading factor of 1.00, reveal factor of 1.00 and static operating mode remain unchanged. The calculator presets “Modern triple glazing” and “Older double glazing” are compared. Only Uw and g differ.

Here, the higher Uw-value does indeed produce a slightly more negative Q_net in the arithmetic annual balance: −279.4 rather than −262.0 kWh/m²a. The reason is the greater solar energy input and, in particular, the larger calculated surplus. The crucial comparison parameter, Q_+, shows the opposite result: the older double-glazed window reaches 168.9 kWh/m²a, whereas the modern triple-glazed window reaches only 51.3 kWh/m²a.

On an hourly basis, the variant with Uw = 2.8 has a lower arithmetic P_net in 2,982 of the 8,760 hours. This occurs under strong irradiation or warm outdoor air. For cooler outdoor conditions, the approximate condition in this comparison is: G_south > 9.74 × (20 − T_outdoor) W/m². In none of these 2,982 hours does the triple-glazed window still have a positive window balance. The additional solar input therefore does not reduce any remaining positive demand; it only increases the surplus.

For these two presets, the answer is therefore: the window with the higher Uw-value can appear better in the arithmetic net balance, but it does not perform better in Q_+, either in hours with a remaining demand or over the year as a whole.

The model is hourly and quasi-steady-state. It assumes a constant indoor temperature and does not represent dynamic room temperatures, thermal mass, ventilation, internal gains, air leakage or a complete building heating-control system.

6. When a Window Can Switch Between Thermal States

A conventional window represents a fixed compromise between U-value, g-value and optical properties, even though sunshine, outdoor temperature and heating or cooling demand change from hour to hour. This raises a design question: must the window remain in the same state at all times?

An adaptive window could allow high solar transmittance when useful sunshine is available in winter, provide greater thermal resistance at night and reduce solar energy input in summer. The operating states alone are therefore insufficient: switching time, control strategy and comfort limits also matter.

FVS is presented here as one possible technical implementation of this adaptive principle. When open, a movable additional thermal protection layer is intended to obstruct daylight and solar input as little as possible; when closed, it is intended to reduce transmission heat flow. Whether this is beneficial in a particular application must be calculated and measured as a function of time.

The fixed compromise between U-value, g-value and solar shading thus becomes a control problem.

Related articles in the Knowledge section: “FVS – How Does Dynamic Thermal Protection Work?”, “Window Retrofit – More Thermal Protection Without Window Replacement” and “From the Component to the Building Energy Balance”.

7. About Calculator 2: “Window Energy Balance”

Calculator 2 illustrates the separate factors affecting a window:

• location, weather data and orientation;

• window area and the proportions of glazing and frame;

• Uw for transmission losses and g for incoming solar energy;

• window reveals, shading and solar shading devices;

• static or switchable additional thermal protection.

From these inputs, the calculator determines P_trans, P_solar,in and the difference between them for every hour. It then sums the hourly values into the Q quantities. This makes it possible to see whether a change primarily affects transmission losses, solar input or the time-dependent balance.

Calculator 2 is an hourly quasi-steady-state comparison model. It illustrates sensitivities and scenarios but is not a substitute for dynamic whole-building simulation, a normative energy assessment or a normative heating-load calculation. ISO/TR 52016-4 remains a methodological reference; no claim of ISO compliance is made for the calculator.

8. Conclusion: A Window Is Not a Single Fixed Value

The U-value remains a key indicator. Without it, transmission heat losses through a window cannot be described reliably. But it is insufficient to assess the window’s overall energy performance.

Such an assessment also requires the g-value, effective glazing area, orientation and shading—and, above all, time. Solar input may be useful in winter yet contribute to overheating in summer. At night, it disappears while transmission continues. Only the time-dependent interaction reveals the window’s effect.

A consistent assessment therefore proceeds as follows:

• U-value: How large are the transmission losses?

• g-value: How much solar energy can enter through the glazing?

• Orientation and shading: How much solar energy is actually available?

• Time of day and season: Is the gain useful or detrimental at that moment?

• Time-dependent balance: What remains over the period being considered?

• Adaptive state: Could the window be operated more effectively under the prevailing conditions?

The balance indicates when an improved static value is sufficient and when a switchable state might be useful.

The U-value describes the transmission-loss side of a component. A time-dependent balance adds solar gains and their usability—without replacing the whole-building energy balance.